JPH1166280A - Medical image processor - Google Patents
Medical image processorInfo
- Publication number
- JPH1166280A JPH1166280A JP9227980A JP22798097A JPH1166280A JP H1166280 A JPH1166280 A JP H1166280A JP 9227980 A JP9227980 A JP 9227980A JP 22798097 A JP22798097 A JP 22798097A JP H1166280 A JPH1166280 A JP H1166280A
- Authority
- JP
- Japan
- Prior art keywords
- image
- intensity
- conversion
- medical image
- signal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000006243 chemical reaction Methods 0.000 claims abstract description 88
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 21
- 238000004458 analytical method Methods 0.000 abstract description 6
- 238000005259 measurement Methods 0.000 abstract description 2
- 238000010276 construction Methods 0.000 abstract 2
- 238000001914 filtration Methods 0.000 description 12
- 238000000034 method Methods 0.000 description 9
- 230000009467 reduction Effects 0.000 description 8
- 230000008859 change Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 5
- 230000008569 process Effects 0.000 description 4
- 230000009466 transformation Effects 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 230000007423 decrease Effects 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 238000003325 tomography Methods 0.000 description 2
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 239000000470 constituent Substances 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000002594 fluoroscopy Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000009206 nuclear medicine Methods 0.000 description 1
Landscapes
- Image Analysis (AREA)
- Apparatus For Radiation Diagnosis (AREA)
- Magnetic Resonance Imaging Apparatus (AREA)
- Image Processing (AREA)
Abstract
Description
【０００１】[0001]
【発明の属する技術分野】この発明は、医用診断機器で
得られるディジタル形式の原医用画像に画質向上用の画
像信号処理を施す医用画像処理装置に係り、特に画質向
上の度合いを高めるための技術に関する。BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a medical image processing apparatus for performing image signal processing for improving image quality on a digital original medical image obtained by a medical diagnostic apparatus, and more particularly to a technique for enhancing the degree of image quality improvement. About.
【０００２】[0002]
【従来の技術】Ｘ線ＣＴ装置（Ｘ線断層撮影装置）で得
られるＸ線ＣＴ画像、あるいは、ＭＲＩ装置（磁気共鳴
断層撮影装置）で得られるＭＲＩ画像などといった医用
診断機器の原医用画像は、各画素ごとに信号濃度値を示
す信号がディジタル信号で構成されているディジタル形
式の画像であるが、画像中にノイズが現れたり、エッジ
（輪郭）が不明瞭であったりして、画質が不十分で的確
な診断が出来ないことが往々にしてある。そこで、医用
診断機器で得られる原医用画像を画像信号処理して、ノ
イズを少なくしたり、エッジ（輪郭）を強調したりし
て、画質を向上させる医用画像処理装置が用いられてい
る。2. Description of the Related Art Original medical images of medical diagnostic equipment such as an X-ray CT image obtained by an X-ray CT apparatus (X-ray tomography apparatus) or an MRI image obtained by an MRI apparatus (magnetic resonance tomography apparatus) are known. A digital image in which a signal indicating a signal density value for each pixel is composed of a digital signal. However, noise appears in the image or edges (contours) are unclear, resulting in poor image quality. It is often the case that an inadequate and inaccurate diagnosis cannot be made. Therefore, a medical image processing apparatus has been used which performs image signal processing on an original medical image obtained by a medical diagnostic device to reduce noise or enhance edges (contours) to improve image quality.
【０００３】従来の医用画像処理装置で実行される画質
向上用の画像信号処理としては、エッジ強調用のラプラ
シアンフィルタリングやハイパスフィルタ、ノイズ低減
用のメディアンフィルタリングやローパスフィルタとい
った空間フィルタリングが挙げられる。ラプラシアンフ
ィルタリングは、例えば注目画素を中心とする３画素×
３画素のサイズの部分画像（領域）に対するディジタル
微分処理である。メディアンフィルタリングは、例えば
注目画素を中心とする３画素×３画素のサイズの部分画
像における中央値選択処理である。ただ、ラプラシアン
フィルタリングの場合、エッジは強調される反面、ノイ
ズも強調されてしまう。逆に、メディアンフィルタリン
グの場合、ノイズは低減される反面、エッジがボケてし
まう。[0003] Image signal processing for improving image quality, which is performed by a conventional medical image processing apparatus, includes spatial filtering such as Laplacian filtering and high-pass filtering for edge enhancement, and median filtering and low-pass filtering for noise reduction. Laplacian filtering is, for example, 3 pixels ×
This is digital differentiation processing for a partial image (region) having a size of three pixels. The median filtering is, for example, a median value selection process in a partial image having a size of 3 pixels × 3 pixels centering on a target pixel. However, in the case of Laplacian filtering, edges are enhanced, but noise is also enhanced. Conversely, in the case of median filtering, noise is reduced, but edges are blurred.
【０００４】そこで、原医用画像の局所領域ごとにノイ
ズ領域（ノイズを含む領域）かエッジ領域（エッジを含
む領域）かといった特徴（領域特徴量）を計測する空間
フィルタリングを先ず行う。この領域特徴量を測る空間
フィルタリングは、例えば局所領域に含まれる画素の信
号強度（濃度値）の分散（バラツキ）を測る手法であ
り、分散が小さな領域は平坦部であってノイズ領域と見
なし、逆に分散が大きな領域は非平坦部であってエッジ
領域と見なす。そして、分散の大きさに応じて、エッジ
強調フィルタとノイズ低減フィルタを組み合わせると、
ある程度、エッジ強調とノイズ低減の両立が図れること
から、画質をそれなりに向上させられる。[0004] Therefore, spatial filtering for measuring a characteristic (region characteristic amount) such as a noise region (a region including a noise) or an edge region (a region including an edge) is first performed for each local region of the original medical image. The spatial filtering for measuring the area feature amount is, for example, a technique of measuring the variance (variation) of the signal intensity (density value) of the pixels included in the local area. An area having a small variance is regarded as a flat area and regarded as a noise area. Conversely, a region having a large variance is a non-flat portion and is regarded as an edge region. Then, when the edge enhancement filter and the noise reduction filter are combined according to the magnitude of the variance,
Since the emphasis of the edge and the noise reduction can be achieved to some extent, the image quality can be improved to some extent.
【０００５】[0005]
【発明が解決しようとする課題】しかしながら、従来の
医用画像処理装置では画質の向上が未だ十分とは言えな
い。ひとつは、局所領域内の画素の信号強度の分散の大
小等の単純な計測量だけでは、濃度の定量把握が出来
ず、濃淡度合いに応じた適切な濃度調整がおこなえない
からである。例えば、濃度差の小さな観察対象エッジと
濃度差の大きな非観察対象エッジとが混在している場
合、濃度の定量把握ができていなければ、両方のエッジ
に対して一律に濃度調整を行うことになる結果、観察対
象エッジは明瞭なエッジとなっても、非観察対象エッジ
の方は強調されすぎてハレーションを引き起こすといっ
た不都合が生じ、両方のエッジに対してそれぞれ適切な
処理をおこあうことは非常に難しいのである。However, the image quality of the conventional medical image processing apparatus has not been sufficiently improved. One is that a simple measurement amount, such as the magnitude of the variance of the signal intensity of the pixels in the local area, cannot be used to quantitatively grasp the density, and that appropriate density adjustment cannot be performed according to the degree of shading. For example, in the case where the observation target edge with a small density difference and the non-observation target edge with a large density difference are mixed, if the density cannot be quantitatively grasped, the density adjustment must be performed uniformly for both edges. As a result, the observed edge becomes a clear edge, but the non-observed edge becomes too emphasized to cause halation, and it is extremely difficult to perform appropriate processing on both edges. It is difficult.
【０００６】もうひとつは、領域特徴量計測用の空間フ
ィルタリングにおける領域サイズを小さくして空間分解
能を高くすると、低周波数情報が十分に計測できず、大
きな周期の濃度差が捉えられない。そうかといって、領
域サイズ（マスクサイズ）を大きくして低周波数情報を
十分に計測しようとすると、領域サイズが大きくなった
分、空間分解能が落ちて、細かな周期の濃度差が捉えら
れないという問題である。実用上は双方の折り合いをつ
けた適当な領域サイズを使用することになるが、必要な
空間分解能と低周波情報が両立できない（折り合いがつ
かない）こともある。On the other hand, if the spatial resolution is increased by reducing the area size in the spatial filtering for measuring the area feature quantity, low frequency information cannot be measured sufficiently, and a large period density difference cannot be captured. On the other hand, if the area size (mask size) is increased and low frequency information is sufficiently measured, the spatial resolution is reduced due to the increase in the area size, and a fine period density difference cannot be captured. That is the problem. In practice, an appropriate area size is used that is a compromise between the two, but the required spatial resolution and low-frequency information may not be compatible (there is no compromise).
【０００７】また、医用画像の場合、医用診断機器の機
種や、撮影部位、あるいは、観察者の個人差によって画
質が大きく変動して、必要な空間分解能や低周波情報が
変化するので、領域サイズを固定して双方の折り合いを
維持することは極めて困難である。[0007] In the case of medical images, the image quality varies greatly depending on the type of medical diagnostic equipment, the imaging region, or individual differences between observers, and the necessary spatial resolution and low-frequency information change. It is extremely difficult to fix the two and maintain the mutual agreement.
【０００８】この発明は、上記問題点に鑑み、医用画像
に対して濃淡度合に応じた適切な濃度調整が行えるとと
もに、画像信号処理の際の高い空間分解能と低周波数情
報の計測とを容易に両立させられて、医用画像の十分な
画質向上が図れる医用画像処理装置を提供することを課
題とする。In view of the above problems, the present invention can perform appropriate density adjustment on a medical image in accordance with the degree of shading, and can easily measure high spatial resolution and low frequency information in image signal processing. It is an object of the present invention to provide a medical image processing apparatus which is compatible with each other and can sufficiently improve the quality of a medical image.
【０００９】[0009]
【課題を解決するための手段】上記課題を解決するため
に、この発明の医用画像処理装置は、医用診断機器で得
られるディジタル形式の原医用画像を記憶する原画像記
憶手段と、原医用画像に画質向上用の画像信号処理を施
す画像処理手段とを備えた医用画像処理装置において、
前記画像処理手段が、原医用画像をウエーブレット変換
して多重解像度分解するウエーブレット変換手段と、原
医用画像における相異なる複数の濃度差に対応付けられ
た多重解像度分解における詳細成分の信号強度のそれぞ
れについて強度調整度を設定する信号強度調整度設定手
段と、前記信号強度調整度設定手段による設定結果に基
づき、詳細成分における必要範囲の各信号強度に対する
調整強度との間の変換関係を示す強度変換特性を求出す
る変換特性求出手段と、前記変換特性求出手段で求出さ
れた強度変換特性に従って各詳細成分に対して強度変換
を実行する強度変換実行手段と、多重解像度分解におけ
る近似成分および強度変換実行済の詳細成分をウエーブ
レット逆変換して画像再構成を実行する画像再構成処理
手段とを具備している。In order to solve the above-mentioned problems, a medical image processing apparatus according to the present invention comprises: an original image storage means for storing an original medical image in a digital format obtained by a medical diagnostic apparatus; Image processing means for performing image signal processing for improving the image quality in the medical image processing apparatus,
The image processing means performs a wavelet transform on the original medical image and performs a multi-resolution decomposition on the original medical image, and a signal intensity of a detailed component in the multi-resolution decomposition associated with a plurality of different density differences in the original medical image. A signal strength adjustment degree setting means for setting the strength adjustment degree for each of the strengths, and an intensity indicating a conversion relationship between the adjustment strength for each signal strength in a necessary range in the detailed component based on a setting result by the signal strength adjustment degree setting means. Conversion characteristic determining means for determining a conversion characteristic, intensity conversion executing means for performing an intensity conversion on each detailed component according to the intensity conversion characteristic determined by the conversion characteristic determining means, approximation in multi-resolution decomposition Image reconstruction processing means for performing wavelet inverse transformation of the component and the detailed component for which the intensity transformation has been performed and performing image reconstruction. That.
【００１０】〔作用〕次に、この発明の医用画像処理装
置によって原医用画像を画像信号処理する際の作用につ
いて説明する。先ず、医用診断機器で得られたディジタ
ル形式の原医用画像が医用画像処理装置の原画像記憶手
段に記憶される。そして、原画像記憶手段に記憶された
原医用画像を、画像処理手段に設けられているウエーブ
レット変換手段でウエーブレット変換して多重解像度分
解し近似成分および詳細成分を得る。画像の場合、２次
元のウエーブレット変換となるので、１個の近似成分と
分解段数×３倍の詳細成分が得られる。ウエーブレット
変換の分解段階数は、必要な低周波情報が得られるよう
に設定すればよく、データ数ｎの場合、最大でｌｏｇ2
ｎとなる。常に最大分解段数を用いても問題はない。[Operation] Next, the operation when the medical image processing apparatus of the present invention processes an original medical image in an image signal will be described. First, an original medical image in digital form obtained by a medical diagnostic device is stored in an original image storage unit of a medical image processing apparatus. Then, the original medical image stored in the original image storage means is subjected to wavelet conversion by the wavelet conversion means provided in the image processing means and subjected to multi-resolution decomposition to obtain approximate components and detailed components. In the case of an image, a two-dimensional wavelet transform is used, so that one approximate component and three-times detailed components are obtained. Decomposition stage number of wavelet transform may be set so that the low frequency information needed to obtain, for a number of data n, a maximum of log 2
n. There is no problem if the maximum number of decomposition stages is always used.
【００１１】このウエーブレット変換では、マスク領域
（窓）というものを設定することなく、位置情報付きの
画像の濃淡関連情報である近似成分および詳細成分が得
られるので、空間フィルタで見られた、領域サイズを固
定することによる空間分解能と低周波情報の競合の心配
はない。また第１段階のウエーブレット変換の詳細成分
が最高周波数情報（最高空間分解能情報）を示し、分解
段階数が１つ上がる度に情報の周波数が半分となる性質
があり（例えば第４段階変換だと最高周波数の１／１６
の低周波数となる）、空間分解能は周波数ごとに常に理
論上可能な限り高く保たれている。画像信号処理過程に
ウエーブレット変換による解析が組み込まれたこの発明
の医用画像処理装置では、低い周波数情報の収集と高い
空間分解能とが容易に理論上可能な限り高い水準で両立
させられることになる。In this wavelet transform, an approximate component and a detailed component, which are shading-related information of an image with position information, can be obtained without setting a mask area (window). There is no concern about competition between spatial resolution and low-frequency information by fixing the region size. In addition, the detailed component of the wavelet transform of the first stage indicates the highest frequency information (highest spatial resolution information), and the frequency of the information is halved every time the number of decomposition stages increases by one (for example, the fourth stage transform And 1/16 of the highest frequency
), The spatial resolution is always kept as high as theoretically possible for each frequency. In the medical image processing apparatus according to the present invention, in which the analysis by the wavelet transform is incorporated in the image signal processing process, the collection of low frequency information and the high spatial resolution can be easily achieved at the highest theoretically possible level. .
【００１２】一方、医用画像の観察者が、画像信号処理
手段の信号強度調整度設定手段により、原医用画像にお
ける相異なる複数の濃度差と対応付けられた多重解像度
分解における詳細成分の信号強度のそれぞれについて強
度調整度（濃度差強調の場合と濃度差低減の場合があ
る）を設定する。つまり、原医用画像における相異なる
複数の濃度差が、多重解像度分解における詳細成分の信
号強度と定量的に対応付けられていて、これを信号強度
調整度設定手段で所望の濃度差が得られるように強度調
整度を設定するのである。そして、この信号強度調整度
設定手段による設定が終わると、この設定結果に基づ
き、変換特性求出手段が直ちに詳細成分における必要範
囲の各信号強度と調整強度との間の変換関係を示す強度
変換特性を求出する。強度調整度を指定した濃度差以外
の濃度差についても、設定した強度調整度に見合った適
当な濃度差に変わるよう必要な強度変換特性が求められ
るのである。On the other hand, the observer of the medical image uses the signal intensity adjustment degree setting means of the image signal processing means to set the signal intensity of the detailed component in the multi-resolution decomposition associated with a plurality of different density differences in the original medical image. The degree of intensity adjustment (the density difference emphasis and the density difference reduction) is set for each. That is, a plurality of different density differences in the original medical image are quantitatively associated with the signal strength of the detailed component in the multi-resolution decomposition, and the desired density difference is obtained by the signal strength adjustment degree setting means. The degree of intensity adjustment is set. When the setting by the signal strength adjustment degree setting means is completed, the conversion characteristic determining means immediately determines the conversion relationship between each signal strength in the required range of the detailed component and the adjustment strength based on the setting result. Find properties. For the density difference other than the density difference for which the intensity adjustment degree is specified, the necessary intensity conversion characteristic is required to change to an appropriate density difference corresponding to the set intensity adjustment degree.
【００１３】続いて、強度変換実行手段により、変換特
性求出手段で求出された強度変換特性に従って各詳細成
分それぞれに対して強度変換が実行される。詳細成分は
濃度差に対応しているので、詳細成分に対する強度変換
は、原医用画像の濃淡度合に応じて適切な濃度調整を行
うことを意味する。詳細成分の強度変換処理に引き続い
て、画像再構成処理手段によって、多重解像度分解にお
ける近似成分および強度変換実行済の詳細成分がウエー
ブレット逆変換されて医用画像の再構成が行われる。こ
のようにして、原医用画像は濃淡度合に応じた適当な濃
度調整が行われ、画質が向上した再構成医用画像が得ら
れる。再構成医用画像は、必要に応じて表示モニタの画
面やフィルムの上に表示されて、観察される。Subsequently, the intensity conversion is executed by the intensity conversion executing means for each of the detailed components in accordance with the intensity conversion characteristics obtained by the conversion characteristic obtaining means. Since the detailed component corresponds to the density difference, the intensity conversion for the detailed component means performing an appropriate density adjustment according to the density of the original medical image. Subsequent to the intensity conversion processing of the detailed component, the approximate component in the multi-resolution decomposition and the detailed component for which the intensity conversion has been performed are inversely wavelet-transformed by the image reconstruction processing unit, and the medical image is reconstructed. In this way, the original medical image is subjected to appropriate density adjustment according to the shading degree, and a reconstructed medical image with improved image quality is obtained. The reconstructed medical image is displayed on a screen of a display monitor or on a film as needed, and is observed.
【００１４】[0014]
【発明の実施の形態】以下、この発明の一実施例を図面
を参照しながら説明する。図１はこの発明の医用画像処
理装置の要部構成を示す全体的ブロック図である。実施
例の医用画像処理装置は、図１に示すように、医用診断
機器１で得られる原医用画像を記憶する原画像メモリ２
と、原医用画像に画質向上用の画像信号処理を施す画像
処理部３と、動作制御用のプログラムに従って装置の各
種動作を制御するコントロール部４を備える他、医用画
像を表示するモニタ５や、コントロール部４への指令入
力や各種のデータ入力をおこなうキーボード６およびマ
ウス７を備えている。以下、各部の構成を具体的に説明
する。An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is an overall block diagram showing a main configuration of a medical image processing apparatus according to the present invention. As shown in FIG. 1, a medical image processing apparatus according to an embodiment includes an original image memory 2 for storing an original medical image obtained by a medical diagnostic apparatus 1.
An image processing unit 3 that performs image signal processing for improving the image quality of the original medical image, a control unit 4 that controls various operations of the apparatus in accordance with an operation control program, a monitor 5 that displays a medical image, A keyboard 6 and a mouse 7 for inputting commands to the control unit 4 and various data are provided. Hereinafter, the configuration of each unit will be specifically described.
【００１５】医用診断機器１としては、Ｘ線ＣＴ装置や
ＭＲＩ装置あるいはＸ線透視撮影装置、ＥＣＴ装置など
の核医学機器などが挙げられ、医用画像として、Ｘ線Ｃ
Ｔ画像やＭＲＩ画像（ＭＲＩ−ＣＴ画像）などの断層画
像の他、Ｘ線透視撮影画像あるいはＲＩ画像などが例示
されるのであるが、もちろんこれら例示のものに限られ
ない。Examples of the medical diagnostic equipment 1 include nuclear medicine equipment such as an X-ray CT apparatus, an MRI apparatus, an X-ray fluoroscopy apparatus, and an ECT apparatus.
In addition to a tomographic image such as a T image or an MRI image (MRI-CT image), an X-ray fluoroscopic image or an RI image is exemplified, but is not limited to these examples.
【００１６】原画像メモリ２は、図２に示すように、例
えば横（Ｘ）方向の番地数が５１２個、縦（Ｙ）方向の
番地数も５１２個の正方形のフレームメモリである。原
医用画像も横（Ｘ）方向の画素数が５１２個、縦（Ｙ）
方向の画素数も５１２個の正方形の画像であって、原画
像メモリ２の各メモリセルに原医用画像の各画素の濃度
値を示すディジタル信号が格納される。As shown in FIG. 2, the original image memory 2 is a square frame memory having, for example, 512 addresses in the horizontal (X) direction and 512 addresses in the vertical (Y) direction. The original medical image also has 512 pixels in the horizontal (X) direction and vertical (Y)
The image is also a square image having 512 pixels in the direction, and a digital signal indicating the density value of each pixel of the original medical image is stored in each memory cell of the original image memory 2.
【００１７】画像処理部３は、図１に示すように、原医
用画像をウエーブレット変換して多重解像度分解するウ
エーブレット変換部１０と、多重解像度分解における近
似成分と詳細成分を記憶するウエーブレット変換結果メ
モリ１１を具備する。ウエーブレット変換は関数ｈを基
底関数として、下記の式（１）において信号を周波数帯
域に次々と所望段階に分解可能な変換手法であり、高周
波から低周波までの所望とする周波数に適合した信号を
得ることができる。また、基底関数は種々あるが、Daub
echies、Symletなどの直交基底を使用すれば、ウエーブ
レット変換で得た近似成分と詳細成分の全てを用いてウ
エーブレット逆変換を行うことで、完全に元の原医用画
像を復元できる特性を有することから原医用画像の情報
が欠落しない。このウエーブレット変換手法は、画像圧
縮技術に応用されたりしており、例えば「ウエーヴレッ
トビギナーズガイド」（榊原進著，１９９５年５月２０
日，東京電機大学出版局発行）に具体的かつ詳細に開示
されている公知の変換手法である。As shown in FIG. 1, the image processing unit 3 includes a wavelet conversion unit 10 for performing wavelet conversion on an original medical image and performing multi-resolution decomposition, and a wavelet for storing approximate components and detailed components in the multi-resolution decomposition. A conversion result memory 11 is provided. The wavelet transform is a transform method that can decompose a signal into desired frequency bands in a frequency band in the following equation (1) using a function h as a basis function, and a signal adapted to a desired frequency from a high frequency to a low frequency. Can be obtained. Although there are various basis functions, Daub
If orthogonal bases such as echies and Symlet are used, the original original medical image can be completely restored by performing the inverse wavelet transform using all of the approximate components and detailed components obtained by the wavelet transform. Therefore, the information of the original medical image is not lost. This wavelet conversion method has been applied to image compression technology, and is described in, for example, "Wavelet Beginner's Guide" (Susumu Sakakibara, May 20, 1995)
This is a well-known conversion method disclosed specifically and in detail by Tokyo Denki University Press.
【００１８】[0018]
【数１】 (Equation 1)
【００１９】但し、 ｆ（ｔ）：任意の波形の信号 Ｗ（ａ，ｂ）：ｆ（ｔ）のウエーブレット変換 ｈ（ａ，ｂ）＝ｈ（ａｔ，ｂ）／√（ａ） ａ：関数の縮率，ｂ：水平軸方向の移動量Where f (t): an arbitrary waveform signal W (a, b): wavelet transform of f (t) h (a, b) = h (at, b) / √ (a) a: Function reduction ratio, b: amount of movement in the horizontal axis direction
【００２０】第１段階変換のウエーブレット変換では、
図３に示す原医用画像（近似成分ＬＬ0 ）Ｐａが分解さ
れて、１個の近似成分ＬＬ-1と３個の詳細成分ＬＨ-1，
ＨＬ -1，ＨＨ-1が得られる。第２段階変換のウエーブレ
ット変換では第１段階変換の近似成分ＬＬ-1がさらに分
解されて、１個の近似成分ＬＬ-2と３個の詳細成分ＬＨ
-2，ＨＬ-2，ＨＨ-2が得られる。以下、同様に第ｊ段階
のウエーブレット変換では、一段階手前の近似成分ＬＬ
-(j-1)が分解されて、近似成分ＬＬ-jと３個の詳細成分
ＬＨ-j，ＨＬ-j，ＨＨ-jが得られる。In the wavelet transformation of the first stage transformation,
The original medical image (approximate component LL) shown in FIG.0) Pa is decomposed
And one approximate component LL-1And three detailed components LH-1,
HL -1, HH-1Is obtained. Waveform of second stage conversion
, The approximate component LL of the first-stage transformation-1But another minute
Solved and one approximation component LL-2And three detailed components LH
-2, HL-2, HH-2Is obtained. Hereinafter, similarly, the j-th stage
In the wavelet transform of, the approximate component LL one step before
-(j-1)Is decomposed into an approximate component LL-jAnd three detailed components
LH-j, HL-j, HH-jIs obtained.
【００２１】近似成分ＬＬ-jはＸ，Ｙ両方向とも所定空
間周波数以下の低周波成分を全て含み、画像的には平均
濃度を示すような画像（平均化処理画像）に当たる。一
方、詳細成分ＬＨ-jはＸ方向が所定空間周波数の高周波
成分であり、詳細成分ＨＬ-jはＹ方向が所定空間周波数
の高周波成分を示あり、詳細成分ＨＨ-jはＸ，Ｙ両方向
が所定空間周波数の高周波成分であって、画像的には濃
度差を示すような画像（微分処理画像）に当たる。The approximate component LL- j includes all low-frequency components below a predetermined spatial frequency in both the X and Y directions, and corresponds to an image (average-processed image) showing an average density. On the other hand, the detailed component LH- j is a high-frequency component having a predetermined spatial frequency in the X direction, the detailed component HL- j is a high-frequency component having a predetermined spatial frequency in the Y direction, and the detailed component HH- j is a component having both the X and Y directions. It is a high-frequency component of a predetermined spatial frequency, and corresponds to an image (differential processing image) that shows a density difference in terms of image.
【００２２】そして、近似成分ＬＬ-jと３個の詳細成分
ＬＨ-j，ＨＬ-j，ＨＨ-jの合計信号個数（合計ピクセル
個数）は、基本的に近似成分ＬＬ-(j-1)の全信号個数
（全ピクセル個数）に等しい。その結果、第１段階から
第ｊ段階までのウエーブレット変換で得られた１個の近
似成分ＬＬ-jと（３×Ｊ）個の全詳細成分の総信号個数
は、基本的に原医用画像の全信号個数に等しい。したが
って、例えば、第１，２段階変換のウエーブレット変換
で得られた１個の近似成分と６個の詳細成分は、図４に
示すように、ウエーブレット変換結果メモリ１１に格納
されるが、ウエーブレット変換結果メモリ１１は基本的
には原画像メモリ２と同じようなメモリセル構成とな
る。また、ウエーブレット変換では段階数が増えると処
理する信号の数が減るので、変換段階が増えても（つま
り、十分な低周波数情報が得られるまで変換段階を増や
しても）演算負荷が肥大化するわけではない。The total signal number (total pixel number) of the approximate component LL- j and the three detailed components LH- j , HL- j , HH- j is basically the approximate component LL- (j-1). Is equal to the total number of signals (the total number of pixels). As a result, the total number of signals of one approximate component LL- j and all (3 × J) detailed components obtained by the wavelet transform from the first stage to the j-th stage is basically equal to the original medical image. Is equal to the total number of signals. Therefore, for example, one approximate component and six detailed components obtained by the wavelet transform of the first and second stage transforms are stored in the wavelet transform result memory 11 as shown in FIG. The wavelet conversion result memory 11 basically has the same memory cell configuration as the original image memory 2. Also, in the wavelet transform, as the number of stages increases, the number of signals to be processed decreases, so even if the number of transform stages increases (ie, the number of transform stages increases until sufficient low-frequency information is obtained), the computational load increases. I do not.
【００２３】一方、画像処理部３は、図１に示すよう
に、原医用画像における相異なる複数（実施例では３
個）の濃度差に対応付けられた多重解像度分解における
詳細成分の信号強度のそれぞれについて強度調整度を設
定する信号強度調整度設定部１２を備えている。この強
度調整度の設定を、具体例に則してＤ明する。図５は対
象の原医用画像（頭部断層像）Ｐａを示し、図６は原医
用画像Ｐａの中央のＸ方向ラインＬｘについての濃度プ
ロファイルを示す。原医用画像Ｐａには、ノイズ（小さ
な凸凹）と弱いエッジ（不明瞭なエッジ）および強いエ
ッジ（明瞭なエッジ）があり、図６に示すように、ノイ
ズは濃度差Ｄ１、弱いエッジは濃度差Ｄ２、強いエッジ
は濃度差Ｄ３である。普通、ノイズは低減する濃度調整
を行い、エッジは強調や現状維持といった濃度調整を行
うことになる。On the other hand, as shown in FIG. 1, the image processing unit 3 has a plurality of different (in the embodiment, 3
And a signal intensity adjustment degree setting unit 12 for setting an intensity adjustment degree for each of the signal intensities of the detailed components in the multi-resolution decomposition associated with the density differences. The setting of the intensity adjustment degree will be described in accordance with a specific example. FIG. 5 shows an original medical image (head tomographic image) Pa of the target, and FIG. 6 shows a density profile of the central X-direction line Lx of the original medical image Pa. The original medical image Pa has noise (small irregularities), weak edges (unclear edges), and strong edges (clear edges). As shown in FIG. 6, noise has a density difference D1 and weak edges have a density difference. D2, the strong edge is the density difference D3. Normally, noise is subjected to density adjustment for reduction, and edges are subjected to density adjustment such as emphasis and maintenance of the current state.
【００２４】図７に示すように、画像における濃度差Ｄ
はウエーブレット変換の多重解像度分解における詳細成
分の信号強度ＷＤと対応関係にある。 ＷＤ＝α×Ｄ＋βAs shown in FIG. 7, the density difference D in the image
Is in correspondence with the signal intensity WD of the detailed component in the multi-resolution decomposition of the wavelet transform. WD = α × D + β
【００２５】なお、原医用画像Ｐａにおける除外対象の
ノイズや観察対象のエッジあるいは非観察対象のエッジ
の各濃度差（注目すべき濃度差）の程度は、医用診断機
器の種類および撮影部位の種類、つまり画像の種類に応
じてほぼ決まっているので、その種類の典型的な原医用
画像を使用し、ノイズやエッジのポイントの濃度差Ｄ１
〜Ｄ３を選択し、３個の詳細成分ＬＨ-j，ＨＬ-j，ＨＨ
-jにおける選択ポイントに対応する位置の各信号強度を
抽出して平均値を求め、これを濃度差Ｄ１〜Ｄ３と対応
付けされた信号強度ＷＤ１〜ＷＤ３として予め信号強度
調整度設定部１２へセットしておく。The degree of each density difference (notable density difference) between the noise to be excluded and the edge of the observation target or the edge of the non-observation target in the original medical image Pa depends on the type of the medical diagnostic equipment and the type of the imaging part. That is, since it is almost determined according to the type of image, a typical original medical image of that type is used, and the density difference D1 between noise and edge points is used.
~ D3, and three detailed components LH- j , HL- j , HH
-j, the respective signal intensities at the positions corresponding to the selected points are extracted and the average value is obtained, and these are set in advance in the signal intensity adjustment degree setting unit 12 as the signal intensities WD1 to WD3 associated with the density differences D1 to D3. Keep it.
【００２６】つまり、信号強度ＷＤ１＝α×｜Ｄ１｜＋
β，ＷＤ２＝α×｜Ｄ２｜＋β，ＷＤ３＝α×｜Ｄ３｜
＋βの演算が行われ、信号強度ＷＤ１〜ＷＤ３が求めら
れて、信号強度調整度設定部１２の内に予めセッされて
いるのである。なお、α，βは、各ウエーブレット変換
の段階数ごとに予め決められる定数である。これらは対
象となるエッジやノイズの濃度変化の空間周波数に依存
する。これらを調整しておくことで、エッジ強調やノイ
ズ低減の対象となる濃度変化の空間周波数を制御でき、
様々な観察目的に対応することができる。したがって、
普通、操作者は、各強度調整度Ａ１〜Ａ３をキーボード
６やマウス７を使って入力するだけで済む。That is, the signal strength WD1 = α × | D1 | +
β, WD2 = α × | D2 | + β, WD3 = α × | D3 |
The calculation of + β is performed, and the signal intensities WD1 to WD3 are obtained and set in the signal intensity adjustment degree setting unit 12 in advance. Note that α and β are constants determined in advance for each number of wavelet transform stages. These depend on the spatial frequency of the edge or noise density change of interest. By adjusting these, you can control the spatial frequency of the density change that is the object of edge enhancement and noise reduction,
It can correspond to various observation purposes. Therefore,
Normally, the operator only needs to input each of the intensity adjustment degrees A1 to A3 using the keyboard 6 and the mouse 7.
【００２７】なお、濃度差Ｄ１〜Ｄ３の選択や信号強度
ＷＤ１〜ＷＤ３のセットは、上のように典型的な医用画
像に基づいて予めセットするのではなくて、実際に画像
信号処理対象とする原医用画像に基づきその都度セット
するようであってもよい。また、各詳細成分ＬＨ-j，Ｈ
Ｌ-j，ＨＨ-jごとに濃度差Ｄ１〜Ｄ３に対応する信号強
度ＷＤ１〜ＷＤ３のセットをそれぞれ行うような構成で
あってもよい。The selection of the density differences D1 to D3 and the setting of the signal intensities WD1 to WD3 are not set in advance on the basis of a typical medical image as described above, but are actually set as image signal processing targets. It may be set each time based on the original medical image. Further, each detailed component LH -j , H
The configuration may be such that the signal intensities WD1 to WD3 corresponding to the density differences D1 to D3 are set for each of L -j and HH -j .
【００２８】さらに、画像処理部３は、信号強度調整度
設定部１２による設定結果である強度調整度Ａ１〜Ａ３
に基づき、詳細成分における必要範囲の各信号強度に対
する調整強度との間の変換関係を示す強度変換特性を求
出する変換特性求出部１３を備えている。濃度差Ｄ１〜
Ｄ３に対応付けされた信号強度ＷＤ１〜ＷＤ３以外の信
号強度も、やはり応分の濃度調整を必要に応じて行うこ
とになるので、変換特性求出部１３が、強度調整度Ａ１
〜Ａ３の間を適当に補完連結し、必要範囲の各信号強度
全体に適用できる強度変換特性ｆ（｜ＷＤｊ｜）を自動
的に求めるのである。強度調整度Ａ１〜Ａ３の間を適当
に補完して強度変換特性ｆ（｜ＷＤｊ｜）が得られるプ
ログラムが制御プログラム中に組み込まれているのであ
る。Further, the image processing unit 3 includes intensity adjustment degrees A1 to A3, which are set results by the signal intensity adjustment degree setting unit 12.
And a conversion characteristic determining unit 13 for determining an intensity conversion characteristic indicating a conversion relationship between the adjustment strength and the signal intensity in the necessary range of the detailed component based on the detailed component. Density difference D1
As for signal intensities other than the signal intensities WD1 to WD3 associated with D3, the corresponding density adjustment is also performed as needed.
To A3 are appropriately complementarily connected to each other, and an intensity conversion characteristic f (│WDj│) applicable to all signal strengths in a necessary range is automatically obtained. A program that appropriately complements the intensity adjustment degrees A1 to A3 to obtain the intensity conversion characteristic f (| WDj |) is incorporated in the control program.
【００２９】強度変換特性ｆ（｜ＷＤｊ｜）が１を越す
と強調調整となり、１未満では低減調整となる。勿論、
強度調整度Ａ１〜Ａ３の設定値に応じて、強度変換特性
ｆ（｜ＷＤｊ｜）のかたちが変化することは言うまでも
ない。例えば、Ａ１＝０，Ａ２＝Ａ３＝１．２と設定さ
れたのであれば、図７に示すように、信号強度ＷＤ１以
下は全て強度調整度Ａ１であるが、信号強度ＷＤ１〜Ｗ
Ｄ２の間は強度調整度Ａ１から強度調整度Ａ２に直線的
に増加し、信号強度ＷＤ２以上は全て強度調整度Ａ２
（＝Ａ３）である強度変換特性ｆａが求出されることに
なる。When the intensity conversion characteristic f (| WDj |) exceeds 1, emphasis adjustment is performed, and when it is less than 1, reduction adjustment is performed. Of course,
It goes without saying that the shape of the intensity conversion characteristic f (| WDj |) changes according to the set values of the intensity adjustment degrees A1 to A3. For example, if A1 = 0 and A2 = A3 = 1.2, as shown in FIG. 7, the signal intensity WD1 and below are all the intensity adjustment degrees A1, but the signal intensities WD1 to W1
During D2, the intensity adjustment degree increases linearly from the intensity adjustment degree A1 to the intensity adjustment degree A2.
The intensity conversion characteristic fa of (= A3) is obtained.
【００３０】あるいは、Ａ１＝０，Ａ２＝１．２，Ａ３
＝１．０と設定されたのであれば、図８に示すように、
信号強度ＷＤ１以下は全て強度調整度Ａ１であるが、信
号強度ＷＤ１〜ＷＤ２の間は強度調整度Ａ１から強度調
整度Ａ２に直線的に増加し、信号強度ＷＤ２〜ＷＤ３の
間は、信号強度ＷＤ３に近くなると強度調整度Ａ３に下
がってゆき、信号強度ＷＤ３以上は全て強度調整度Ａ３
である強度変換特性ｆｂが求出されることになる。強度
変換特性ｆｂでは、強いエッジの場合、さらに強調する
と不必要な感じとなり、アーティファクトの発生をみた
りすることから、信号強度が一定値を越すような範囲は
強調せずに現状維持とする調整が行われることになる。Alternatively, A1 = 0, A2 = 1.2, A3
= 1.0, as shown in FIG.
The signal intensity WD1 and below are all the intensity adjustment A1, but linearly increase from the intensity adjustment A1 to the intensity adjustment A2 during the signal intensity WD1 to WD2, and the signal intensity WD3 between the signal intensity WD2 and WD3. , The intensity adjustment level decreases to the intensity adjustment degree A3, and the signal intensity WD3 and higher are all adjusted to the intensity adjustment degree A3.
Is obtained. In the intensity conversion characteristic fb, in the case of a strong edge, further emphasis becomes unnecessary, and an occurrence of an artifact is observed. Therefore, the range in which the signal intensity exceeds a certain value is not adjusted and the current state is maintained without being emphasized. Will be performed.
【００３１】なお、詳細成分ＬＨ-j，ＨＬ-j，ＨＨ-jご
とに濃度差Ｄ１〜Ｄ３に対応する信号強度ＷＤ１〜ＷＤ
３のセットがそれぞれ行われる構成の場合、強度変換特
性（｜ＷＤｊ｜）も、細成分ＬＨ-j，ＨＬ-j，ＨＨ-jご
とに求出されるので、強度変換特性（｜ＷＤｊ｜）の求
出数が３倍になる。The signal intensities WD1 to WD corresponding to the density differences D1 to D3 for each of the detailed components LH- j , HL- j , and HH- j.
3, the intensity conversion characteristics (| WDj |) are also found for each of the fine components LH -j , HL -j , and HH -j , so that the intensity conversion characteristics (| WDj |) Is tripled.
【００３２】さらに、画像処理部３は、強度変換特性ｆ
（｜ＷＤｊ｜）に従って各詳細成分に対して強度変換を
実行する強度変換実行部１４と、多重解像度分解におけ
る近似成分および強度変換実行済の詳細成分をウエーブ
レット逆変換して画像再構成をおこなう画像再構成処理
部１５も備えている。強度変換実行部１４では、各詳細
成分の全信号強度ＤＷｊの各々について対応するＡｊ
（＝ｆ（｜ＷＤｊ｜）が求められた後、Ａｊ×ＤＷｊの
演算が行われ、調整済の信号強度Ｗｄｊ（＝Ａｊ×ＤＷ
ｊ）が求められる。求められた信号強度Ｗｄｊは、ウエ
ーブレット変換結果メモリ１１の対応するメモリセルの
信号強度ＷＤｊと置き換えられることになる。画像再構
成処理部１５は、強度変換済の全詳細成分と（強度変換
はされていない）近似成分とをウエーブレット逆変換し
て画像再構成を行い、再構成医用画像を得た後、再構成
医用画像を表示するためのモニタ５の方へ送出する機能
を有する。Further, the image processing section 3 has an intensity conversion characteristic f
(| WDj |) performs an intensity conversion on each of the detailed components, and performs an inverse wavelet transform on the approximate components and the detailed components on which the intensity conversion has been performed in the multi-resolution decomposition to perform image reconstruction. An image reconstruction processing unit 15 is also provided. In the intensity conversion execution unit 14, the corresponding Aj for each of the total signal intensities DWj of the respective detailed components
(= F (| WDj |) is calculated, Aj × DWj is calculated, and the adjusted signal strength Wdj (= Aj × DW
j) is required. The obtained signal strength Wdj is replaced with the signal strength WDj of the corresponding memory cell of the wavelet conversion result memory 11. The image reconstruction processing unit 15 performs an image reconstruction by performing a wavelet inverse transform on all the detailed components subjected to the intensity conversion and the approximate components (not subjected to the intensity conversion) to obtain a reconstructed medical image. It has a function of transmitting the constituent medical images to the monitor 5 for displaying.
【００３３】なお、画像処理部３におけるウエーブレッ
ト変換部１０、信号強度調整度設定部１２、変換特性求
出部１３、強度変換実行部１４、および、画像再構成処
理部１５は、コンピュータ（ＣＰＵ）およびその制御プ
ログラムなどを中心に構成されるものである。It should be noted that the wavelet converter 10, the signal intensity adjustment degree setting unit 12, the conversion characteristic calculating unit 13, the intensity conversion executing unit 14, and the image reconstruction processing unit 15 in the image processing unit 3 are each composed of a computer (CPU ) And its control program.
【００３４】続いて、以上に説明した実施例の医用画像
処理装置により図５の原医用画像Ｐａを画像信号処理す
る際の装置動作を、画像信号処理動作の流れを示すフロ
ーチャートである図１１に則して説明する。前述のよう
に、図５に示す原医用画像Ｐａには、ノイズと弱いエッ
ジ（不明瞭なエッジ）および強いエッジ（明瞭なエッ
ジ）がある。ノイズの濃度差Ｄ１、弱いエッジの濃度差
Ｄ２、強いエッジの濃度差Ｄ３は、図６の濃度プロファ
イルに示すように、濃度差Ｄ１＜濃度差Ｄ２＜濃度差Ｄ
３の関係にある。以下の画像信号処理においては、ノイ
ズは低減、弱いエッジは明瞭化し、強いエッジは現状を
維持する濃度調整をおこなうことになる。FIG. 11 is a flowchart showing the flow of the image signal processing operation when the medical image processing apparatus of the above-described embodiment performs image signal processing on the original medical image Pa in FIG. It will be explained in general. As described above, the original medical image Pa shown in FIG. 5 includes noise and weak edges (unclear edges) and strong edges (clear edges). The density difference D1 of the noise, the density difference D2 of the weak edge, and the density difference D3 of the strong edge are, as shown in the density profile of FIG. 6, the density difference D1 <the density difference D2 <the density difference D.
There is a relationship of 3. In the following image signal processing, noise is reduced, weak edges are clarified, and strong edges are subjected to density adjustment maintaining the current state.
【００３５】〔ステップＦ１〕先ず、医用診断機器１か
ら原医用画像Ｐａが送出されて原画像メモリ２に記憶さ
れる。必要に応じて、原画像メモリ２に記憶された原医
用画像Ｐａはモニタ５の画面に映し出され、操作者など
の観察に供される。[Step F1] First, an original medical image Pa is transmitted from the medical diagnostic apparatus 1 and stored in the original image memory 2. If necessary, the original medical image Pa stored in the original image memory 2 is displayed on the screen of the monitor 5, and is used for observation by an operator or the like.
【００３６】〔ステップＦ２〕 画像処理部３のウエー
ブレット変換部１０が、原画像メモリ２の原医用画像を
第４段階までウエーブレット変換して多重解像度分解
し、１個の近似成分と１２個の詳細成分が得られる。得
られた近似成分および詳細成分はウエーブレット変換結
果メモリ１１に送られて格納される。なお、図８の濃度
プロファイルに対応する詳細成分の信号強度プロファイ
ル（濃度差プロファイル）を図９に示す。信号強度ＷＤ
１がノイズの濃度差Ｄ１、信号強度ＷＤ２が弱いエッジ
の濃度差Ｄ２、信号強度ＷＤ３が強いエッジの濃度差Ｄ
３にそれぞれ対応している。[Step F2] The wavelet transform unit 10 of the image processing unit 3 performs wavelet transform on the original medical image in the original image memory 2 to the fourth stage and decomposes the image into multiple resolutions. Are obtained. The obtained approximate component and detailed component are sent to the wavelet conversion result memory 11 and stored. FIG. 9 shows a signal intensity profile (density difference profile) of a detailed component corresponding to the density profile of FIG. Signal strength WD
1 is the density difference D1 of the noise, the density difference D2 of the edge having a weak signal strength WD2, and the density difference D of the edge having a strong signal strength WD3.
3 respectively.
【００３７】〔ステップＦ３〕操作者が、信号強度調整
度設定部１２に設定する強度調整度Ａ１〜Ａ３の設定値
として、Ａ１＝０，Ａ２＝１．２，Ａ３＝１．０をキー
ボート６から入力する。ノイズの濃度差は０とし、弱い
エッジの濃度差は１．２倍に強調し、強いエッジの濃度
差はそのままとする設定である。[Step F3] The operator sets A1 = 0, A2 = 1.2, and A3 = 1.0 as keyboard setting values of the intensity adjustment degrees A1 to A3 set in the signal intensity adjustment degree setting unit 12. Enter from. The density difference of noise is set to 0, the density difference of weak edges is emphasized by 1.2 times, and the density difference of strong edges is kept as it is.
【００３８】〔ステップＦ４〕設定された強度調整度Ａ
１〜Ａ３に基づき、変換特性求出部１３が強度変換特性
を求出する。今はＡ１＝０，Ａ２＝１．２，Ａ３＝１．
０であるから、図８に示す強度変換特性ｆｂが求出され
る。[Step F4] Set degree of intensity adjustment A
Based on 1 to A3, the conversion characteristic determining unit 13 determines the intensity conversion characteristic. Now A1 = 0, A2 = 1.2, A3 = 1.
Since it is 0, the intensity conversion characteristic fb shown in FIG. 8 is obtained.
【００３９】〔ステップＦ５〕図８の強度変換特性ｆｂ
に従って、強度変換実行部１４が、全詳細成分の各信号
強度ＷＤｊに対し、Ａｊ×ＷＤｊ＝Ｗｄｊの演算を実行
し、ウエーブレット変換結果メモリ１１の対応するメモ
リセルの信号強度ＷＤｊから強度変換済の信号強度Ｗｄ
ｊに全て置き換える。[Step F5] The intensity conversion characteristic fb shown in FIG.
, The intensity conversion execution unit 14 performs an operation of Aj × WDj = Wdj for each signal intensity WDj of all the detailed components, and the intensity conversion is performed from the signal intensity WDj of the corresponding memory cell of the wavelet conversion result memory 11. Signal strength Wd
Replace all with j.
【００４０】なお、図９の信号強度プロファイルの強度
変換後の状態を図１０に示す。図１０に示すように、強
度変換特性ｆｂに従う強度変換により、詳細成分の信号
強度プロファイルからはノイズに相当する信号がうまく
消えている（ノイズの低減が実現されている）。また、
弱いエッジを表す成分は１．２倍に、強いエッジを表す
成分は等倍保存されている。FIG. 10 shows a state of the signal intensity profile of FIG. 9 after the intensity conversion. As shown in FIG. 10, the signal corresponding to the noise has been successfully eliminated from the signal intensity profile of the detailed component by the intensity conversion according to the intensity conversion characteristic fb (the noise has been reduced). Also,
A component representing a weak edge is preserved at 1.2 times, and a component representing a strong edge is preserved at the same magnification.
【００４１】〔ステップＦ６〕 信号強度ＷＤｊから信
号強度Ｗｄｊへの置き換え完了に続き、画像再構成処理
部１５が、強度変換済の１２個の詳細成分および（強度
変換はされていない）近似成分をウエーブレット逆変換
して画像再構成を行い、再構成医用画像を得る。[Step F6] Following the completion of the replacement of the signal strength WDj with the signal strength Wdj, the image reconstruction processing unit 15 converts the twelve detailed components subjected to the intensity conversion and the approximate components (not subjected to the intensity conversion) into the 12 detailed components. Image reconstruction is performed by inverse wavelet transform to obtain a reconstructed medical image.
【００４２】〔ステップＦ７〕 モニタ５の画面には、
ノイズは消え、不明瞭だった弱いエッジが強調された十
分な画質の再構成医用画像か映し出される。[Step F7] On the screen of the monitor 5,
The noise disappears and appears as a reconstructed medical image of sufficient quality with the weak edges that were obscured enhanced.
【００４３】このように、実施例装置の場合、原医用画
像Ｐａにおけるノイズやエッジの濃度差の量に応じてそ
れぞれ適切な濃度調整が行われる。それに、原医用画像
の信号処理過程でマスク領域を要しないウエーブレット
変換で原医用画像Ｐａの濃度解析が行われる。ウエーブ
レット変換は、高周波域から十分な低周波域までの濃度
解析結果をもたらすので、画像信号処理の際の高い空間
分解能と十分な低周波数情報とが容易に確保される。し
たがって、再構成された医用画像は十分に画質が向上し
たものとなる。As described above, in the case of the apparatus according to the embodiment, appropriate density adjustment is performed in accordance with the amount of noise and the density difference between edges in the original medical image Pa. In addition, the density analysis of the original medical image Pa is performed by a wavelet transform that does not require a mask area in the signal processing process of the original medical image. Since the wavelet transform provides a result of density analysis from a high frequency range to a sufficiently low frequency range, a high spatial resolution and sufficient low frequency information at the time of image signal processing are easily secured. Therefore, the reconstructed medical image has a sufficiently improved image quality.
【００４４】この発明は上記の実施例に限られるもので
はなく、以下のように変形実施することもできる。 （１）上記の実施例では、ウエーブレット変換が４段階
変換であったが、変換段階数が高いほど低周波情報が十
分となり、演算量はそれほど増大しないので、分解段階
数は十分大きく取るのが好ましい。The present invention is not limited to the above embodiment, but can be modified as follows. (1) In the above embodiment, the wavelet transform is a four-stage transform. However, the higher the number of transform stages, the more low-frequency information is sufficient and the amount of calculation does not increase so much. Is preferred.
【００４５】（２）上記の実施例では、原医用画像Ｐａ
における３個の濃度差のそれぞれについて強度調整度Ａ
１〜Ａ３を設定する構成であったが、原医用画像Ｐａに
おける２個あるいは４個以上の濃度差のそれぞれについ
て強度調整度を設定する構成のものが、変形例として挙
げられる。(2) In the above embodiment, the original medical image Pa
Intensity adjustment A for each of the three density differences in
Although a configuration in which 1 to A3 are set, a configuration in which the intensity adjustment degree is set for each of two or four or more density differences in the original medical image Pa is given as a modified example.
【００４６】（３）上記の実施例は強度調整度Ａ１〜Ａ
３をキーボード等から入力する構成であったが、予めＡ
１＝０．４、Ａ２＝１．５、Ａ３＝１．０等の適切な結
果が期待できる値を準備しておいてもよい。(3) In the above embodiment, the strength adjustment degrees A1 to A
3 was input from a keyboard or the like.
A value such as 1 = 0.4, A2 = 1.5, A3 = 1.0, etc., for which an appropriate result can be expected may be prepared.
【００４７】（４）上記の実施例では詳細成分ＷＤその
ものについて、濃度差Ｄとの関連付け（ＷＤ＝αＤ×
β）と強度変換特性ｆ（｜ＷＤ｜）の求出を行ったが、
各詳細成分ＬＨ-j，ＨＬ-j，ＨＨ-jの相互相関係数ＣＯ
ＲＲに対して濃度差Ｄとの関連付けＣＯＲＲ＝αＤ＋β
と強度変換特性ｆ（ＣＯＲＲ）の求出を行ってもよい。
ここで、相互相関係数ＣＯＲＲは、例えば、 ＣＯＲＲj [y][x]＝(1/3) （｜ＬＨ-j[y][x]｜＋｜ＨＬ
-j[y][x]｜＋｜ＨＨ-j[y][x]｜） である。これは方向に依らず、[y][x]点における分解段
数ｊに相当する空間周波数の濃度変化の大きさを表す量
となる。また、例えば、 ＣＯＲＲj [y][x]＝(1/2) （｜ＬＨ-j[y][x]｜＋｜ＬＨ
-j+1[y/2][x/2]｜） とすれば、分解段数ｊとｊ−１の空間周波数の相互相関
を表す量となる。(4) In the above embodiment, the detailed component WD itself is associated with the density difference D (WD = αD ×
β) and the intensity conversion characteristic f (| WD |)
The cross-correlation coefficient CO of each detailed component LH- j , HL- j , HH- j
Correlation of RR with concentration difference D CORR = αD + β
And the strength conversion characteristic f (CORR) may be determined.
Here, the cross-correlation coefficient CORR is, for example, CORR j [y] [x] = (1/3) (| LH −j [y] [x] | + | HL
-j [y] [x] | + | HH- j [y] [x] |). This is an amount that indicates the magnitude of the change in the density of the spatial frequency corresponding to the number of decomposition steps j at the point [y] [x], regardless of the direction. Also, for example, CORR j [y] [x] = (1/2) (| LH -j [y] [x] | + | LH
-j + 1 [y / 2] [x / 2] |) is a quantity representing the cross-correlation between the spatial frequency of the decomposition stage j and j-1.
【００４８】[0048]
【発明の効果】以上の説明から明らかなように、この発
明の医用画像処理装置によれば、原医用画像におけるノ
イズやエッジなど注目すべき各濃度差の量に応じてそれ
ぞれ適切な濃度調整が行われる上、ウエーブレット変換
により原医用画像の位置情報付き濃度解析が行われるの
で、高周波域から十分な低周波域まで濃淡差解析結果を
もたらすことから、画像信号処理の際の高い空間分解能
と十分な低周波数情報とを確保することができ、医用画
像の十分な画質向上が図れる。As is apparent from the above description, according to the medical image processing apparatus of the present invention, appropriate density adjustment can be performed in accordance with the amount of each density difference to be noted, such as noise and edges, in the original medical image. In addition to this, the density analysis with the position information of the original medical image is performed by the wavelet transform, so that the density difference analysis result is obtained from a high frequency range to a sufficiently low frequency range. Sufficient low-frequency information can be ensured, and the image quality of medical images can be sufficiently improved.
【図１】実施例の医用画像処理装置の要部構成を示す全
体的ブロック図である。FIG. 1 is an overall block diagram illustrating a main configuration of a medical image processing apparatus according to an embodiment.
【図２】実施例装置の原画像メモリのアドレス構成を示
す模式図である。FIG. 2 is a schematic diagram showing an address configuration of an original image memory of the embodiment device.
【図３】医用画像のウエーブレット変換結果を説明する
模式図である。FIG. 3 is a schematic diagram illustrating a wavelet conversion result of a medical image.
【図４】実施例でのウエーブレット変換結果メモリの記
憶内容例を示す模式図である。FIG. 4 is a schematic diagram showing an example of stored contents of a wavelet conversion result memory in the embodiment.
【図５】濃度調整対象の原医用画像例を示す平面図であ
る。FIG. 5 is a plan view showing an example of an original medical image to be adjusted in density.
【図６】原医用画像の中央のＸ方向ラインについての濃
度プロファイルを示すグラフである。FIG. 6 is a graph showing a density profile of a central X-direction line of an original medical image.
【図７】実施例で求出された強度変換特性を示すグラフ
である。FIG. 7 is a graph showing the intensity conversion characteristics determined in the example.
【図８】実施例で求出された他の強度変換特性を示すグ
ラフである。FIG. 8 is a graph showing another intensity conversion characteristic obtained in the example.
【図９】ウエーブレット変換による詳細成分の信号強度
の強度変換前の状態を示すグラフである。FIG. 9 is a graph showing a state before signal intensity conversion of a signal intensity of a detailed component by wavelet conversion.
【図１０】ウエーブレット変換による詳細成分の信号強
度の強度変換後の状態を示すグラフである。FIG. 10 is a graph showing a state after signal strength conversion of a signal strength of a detailed component by wavelet transform.
【図１１】実施例装置による画像信号処理動作の流れを
示すフローチャートである。FIG. 11 is a flowchart illustrating a flow of an image signal processing operation performed by the apparatus according to the embodiment.
１…医用診断機器 ２…原画像メモリ ３…画像処理部 １０…ウエーブレット変換部 １２…信号強度調整度設定部 １３…変換特性求出部 １４…強度変換実行部 １５…画像再構成処理部 Ｄ１〜Ｄ３…濃度差 Ａ１〜Ａ３…強度調整度 ＷＤ１〜ＷＤ３…詳細成分の信号強度 ｆａ，ｆｂ…強度変換特性 DESCRIPTION OF SYMBOLS 1 ... Medical diagnostic equipment 2 ... Original image memory 3 ... Image processing part 10 ... Wavelet conversion part 12 ... Signal strength adjustment degree setting part 13 ... Conversion characteristic calculation part 14 ... Intensity conversion execution part 15 ... Image reconstruction processing part D1 ＤD3: density difference A1 to A3: intensity adjustment degree WD1 to WD3: signal intensity of detailed component fa, fb: intensity conversion characteristics
Claims (1)
の原医用画像を記憶する原画像記憶手段と、原医用画像
に画質向上用の画像信号処理を施す画像処理手段とを備
えた医用画像処理装置において、前記画像処理手段が、
原医用画像をウエーブレット変換して多重解像度分解す
るウエーブレット変換手段と、原医用画像における相異
なる複数の濃度差に対応付けられた多重解像度分解にお
ける詳細成分の信号強度のそれぞれについて強度調整度
を設定する信号強度調整度設定手段と、前記信号強度調
整度設定手段による設定結果に基づき、詳細成分におけ
る必要範囲の各信号強度に対する調整強度との間の変換
関係を示す強度変換特性を求出する変換特性求出手段
と、前記変換特性求出手段で求出された強度変換特性に
従って各詳細成分に対して強度変換を実行する強度変換
実行手段と、多重解像度分解における近似成分および強
度変換実行済の詳細成分をウエーブレット逆変換して画
像再構成を実行する画像再構成処理手段とを具備してい
ることを特徴とする医用画像処理装置。1. A medical image processing apparatus comprising: an original image storage unit for storing an original medical image in a digital format obtained by a medical diagnostic apparatus; and an image processing unit for performing image signal processing for improving the image quality of the original medical image. In the image processing means,
A wavelet transform unit that performs wavelet transform on the original medical image and performs multi-resolution decomposition, and an intensity adjustment degree for each of the signal intensities of the detailed components in the multi-resolution decomposition corresponding to a plurality of different density differences in the original medical image. Based on a setting result of the signal strength adjustment degree setting means to be set and the signal strength adjustment degree setting means, an intensity conversion characteristic indicating a conversion relationship between the adjustment strength for each signal strength in a required range of the detailed component is obtained. Conversion characteristic determining means, intensity conversion executing means for performing intensity conversion on each detailed component in accordance with the intensity conversion characteristic determined by the conversion characteristic determining means, approximate component and intensity conversion executed in multi-resolution decomposition Image reconstruction processing means for performing an image reconstruction by performing a wavelet inverse transform on the detailed component of Use image processing apparatus.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP9227980A JPH1166280A (en) | 1997-08-25 | 1997-08-25 | Medical image processor |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP9227980A JPH1166280A (en) | 1997-08-25 | 1997-08-25 | Medical image processor |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| JPH1166280A true JPH1166280A (en) | 1999-03-09 |
Family
ID=16869283
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| JP9227980A Pending JPH1166280A (en) | 1997-08-25 | 1997-08-25 | Medical image processor |
Country Status (1)
| Country | Link |
|---|---|
| JP (1) | JPH1166280A (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2001307066A (en) * | 2000-04-21 | 2001-11-02 | Matsushita Electric Ind Co Ltd | Cell image analyzer and cell image analyzing method |
| WO2002045020A1 (en) * | 2000-11-30 | 2002-06-06 | Canon Kabushiki Kaisha | Image processing device, image processing method, storage medium, and program |
| WO2002045019A1 (en) * | 2000-11-30 | 2002-06-06 | Canon Kabushiki Kaisha | Image processing device, image processing method, recording medium and program |
| JP2002358518A (en) * | 2001-06-01 | 2002-12-13 | Canon Inc | Device, system and method for processing image, storage medium and program |
| JP2003139678A (en) * | 2001-10-30 | 2003-05-14 | Matsushita Electric Ind Co Ltd | Cell image analyzer, cell image analyzing method and recording medium |
| WO2005004057A1 (en) * | 2003-07-04 | 2005-01-13 | Japan Science And Technology Agency | Image processor, image processing method, program and recording medium |
| JP2005109867A (en) * | 2003-09-30 | 2005-04-21 | Konica Minolta Medical & Graphic Inc | Image processing method, image processor and image processing program |
| JP2006127420A (en) * | 2004-11-01 | 2006-05-18 | Sony Corp | Image processor, image processing method, and program |
| JP2008033962A (en) * | 2000-11-09 | 2008-02-14 | Canon Inc | Image processor, its method, program and storage medium |
| JP2010244360A (en) * | 2009-04-07 | 2010-10-28 | Canon Inc | Image processing apparatus, image processing method, and computer program |
| JP2013025368A (en) * | 2011-07-15 | 2013-02-04 | Tokyo Institute Of Technology | Noise level estimation device, noise level estimation method and program |
-
1997
- 1997-08-25 JP JP9227980A patent/JPH1166280A/en active Pending
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2001307066A (en) * | 2000-04-21 | 2001-11-02 | Matsushita Electric Ind Co Ltd | Cell image analyzer and cell image analyzing method |
| JP2008033962A (en) * | 2000-11-09 | 2008-02-14 | Canon Inc | Image processor, its method, program and storage medium |
| JP4630893B2 (en) * | 2000-11-09 | 2011-02-09 | キヤノン株式会社 | Image processing apparatus, method thereof, program, and storage medium |
| US7447376B2 (en) | 2000-11-30 | 2008-11-04 | Canon Kabushiki Kaisha | Image processing apparatus, image processing method, storage medium, and program |
| US7558434B2 (en) | 2000-11-30 | 2009-07-07 | Canon Kabushiki Kaisha | Image processing apparatus, image processing method, storage medium, and program |
| WO2002045020A1 (en) * | 2000-11-30 | 2002-06-06 | Canon Kabushiki Kaisha | Image processing device, image processing method, storage medium, and program |
| US7561750B2 (en) | 2000-11-30 | 2009-07-14 | Canon Kabushiki Kaisha | Image processing apparatus, image processing method, storage medium, and program |
| WO2002045019A1 (en) * | 2000-11-30 | 2002-06-06 | Canon Kabushiki Kaisha | Image processing device, image processing method, recording medium and program |
| US7076111B2 (en) | 2000-11-30 | 2006-07-11 | Canon Kabushiki Kaisha | Image processing apparatus, image processing method, storage medium, and program |
| US7079700B2 (en) | 2000-11-30 | 2006-07-18 | Canon Kabushiki Kaisha | Image processing apparatus, image processing method, storage medium, and program |
| US7248748B2 (en) | 2000-11-30 | 2007-07-24 | Canon Kabushiki Kaisha | Image processing apparatus, image processing method, storage medium, and program |
| JP2002358518A (en) * | 2001-06-01 | 2002-12-13 | Canon Inc | Device, system and method for processing image, storage medium and program |
| JP2003139678A (en) * | 2001-10-30 | 2003-05-14 | Matsushita Electric Ind Co Ltd | Cell image analyzer, cell image analyzing method and recording medium |
| WO2005004057A1 (en) * | 2003-07-04 | 2005-01-13 | Japan Science And Technology Agency | Image processor, image processing method, program and recording medium |
| JP2005109867A (en) * | 2003-09-30 | 2005-04-21 | Konica Minolta Medical & Graphic Inc | Image processing method, image processor and image processing program |
| JP2006127420A (en) * | 2004-11-01 | 2006-05-18 | Sony Corp | Image processor, image processing method, and program |
| JP4656383B2 (en) * | 2004-11-01 | 2011-03-23 | ソニー株式会社 | Image processing apparatus, image processing method, and program |
| JP2010244360A (en) * | 2009-04-07 | 2010-10-28 | Canon Inc | Image processing apparatus, image processing method, and computer program |
| JP2013025368A (en) * | 2011-07-15 | 2013-02-04 | Tokyo Institute Of Technology | Noise level estimation device, noise level estimation method and program |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US5150421A (en) | System for automated transformation of gray level of image | |
| CN103961130B (en) | So that C-arm system adapts to the method to provide three-dimensional imaging information | |
| US9245323B2 (en) | Medical diagnostic device and method of improving image quality of medical diagnostic device | |
| US7218763B2 (en) | Method for automated window-level settings for magnetic resonance images | |
| US11727609B2 (en) | Limited-angle CT reconstruction method based on anisotropic total variation | |
| US20100286525A1 (en) | Ultrasound diagnosis apparatus | |
| US7020319B2 (en) | Method and apparatus for generating three-dimensional, multiply resolved volume images of an examination subject | |
| US20020196907A1 (en) | Image processing apparatus, image processing system, image processing method, program, and storage medium | |
| JPH11155834A (en) | Medial imaging system and method for forming medical image | |
| JPH1031745A (en) | Automatic image analysis method | |
| JP2012524329A (en) | Multiscale image normalization and enhancement | |
| JP2000516796A (en) | Image noise compression method | |
| CN111065335B (en) | Medical image processing apparatus and medical image processing method | |
| US9619893B2 (en) | Body motion detection device and method | |
| CN106530236B (en) | Medical image processing method and system | |
| US20040105528A1 (en) | Method and system for tomosynthesis image enhancement using transverse filtering | |
| JPH1166280A (en) | Medical image processor | |
| WO2012033199A1 (en) | Ultrasound diagnostic devic, medical image-processing device and medical image-processing method | |
| JP2016214857A (en) | Medical image processor and medical image processing method | |
| US20030081820A1 (en) | Method for contrast matching of multiple images of the same object or scene to a common reference image | |
| Mekle et al. | Evaluation of a multiscale enhancement protocol for digital mammography | |
| JP4679095B2 (en) | Image processing apparatus, image processing method, and program | |
| JP2008142178A (en) | Radiological image processing apparatus | |
| Fonseca et al. | X-ray image enhancement: A technique combination approach | |
| JP2009054013A (en) | Image processor |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| A621 | Written request for application examination |
Free format text: JAPANESE INTERMEDIATE CODE: A621 Effective date: 20040406 |
|
| A977 | Report on retrieval |
Free format text: JAPANESE INTERMEDIATE CODE: A971007 Effective date: 20070420 |
|
| A131 | Notification of reasons for refusal |
Free format text: JAPANESE INTERMEDIATE CODE: A131 Effective date: 20070703 |
|
| A521 | Request for written amendment filed |
Free format text: JAPANESE INTERMEDIATE CODE: A523 Effective date: 20070903 |
|
| A131 | Notification of reasons for refusal |
Free format text: JAPANESE INTERMEDIATE CODE: A131 Effective date: 20071120 |
|
| A02 | Decision of refusal |
Free format text: JAPANESE INTERMEDIATE CODE: A02 Effective date: 20080408 |