Recent advances and future perspectives of machine learning techniques offer promising applications in medical imaging. Machine learning has the potential to improve different steps of the radiology workflow including order scheduling and triage, clinical decision support systems, detection and interpretation of findings, postprocessing and dose estimation, examination quality control, and radiology reporting. In this article, the authors review examples of current applications of machine learning and artificial intelligence techniques in diagnostic radiology. In addition, the future impact and natural extension of these techniques in radiology practice are discussed.

© RSNA, 2018

https://pubs.rsna.org/doi/pdf/10.1148/radiol.2018171820

Current Applications and Future Impact of Machine Learning in Radiology.

With the increase in breast cancer risk over the years, there are many factors estimated that lead to it. However, till date which factor is majorly involved in development of breast cancer or which factor accounts more is not clearly evident. Mammography technique accounting for 80-90% of cancer being detected is believed to be the best method of detection. While mammographic density is manifested by increased proliferation of fat, stoma, epithelium and connective tissue, it is considered to be a risk factor for development of breast cancer. The current study was thus conducted to find out whether the mammographic density is actually a risk factor for development of breast cancer and to find out the better detection tool available. For this, the methodology adopted was review of various journals and studies already published with respect to mammographic density and its risk on development of breast cancer. The conclusion of the current study as well as from another comparable study was that the frequency of screening might be dependent on breast density and in such cases diagnostic techniques such as “digital mammography, ultra sonography and magnetic resonance imaging” may prove to be better detection tools. Moreover, recent studies have also suggested that mammographic density as a marker for risk of developing breast cancer holds true however, this fact needs to be evaluated further. Article DOI:  https://dx.doi.org/10.20319/lijhls.2016.22.4854 This work is licensed under the Creative Commons Attribution-Non-commercial 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc/4.0/ or send a letter to Creative Commons, PO Box 1866, Mountain View, CA 94042, USA.

Mammographic density and the risk and detection of breast cancer

Medical breast ultrasound image segmentation by machine learning.

Breast cancer is among world's second most occurring cancer in all types of cancer. Most common cancer among women worldwide is breast cancer. There is always need of advancement when it comes to medical imaging. Early detection of cancer followed by the proper treatment can reduce the risk of deaths. Machine learning can help medical professionals to diagnose the disease with more accuracy. Where deep learning or neural networks is one of the techniques which can be used for the classification of normal and abnormal breast detection. CNN can be used for this detection. Mammograms-MIAS dataset is used for this purpose, having 322 mammograms in which almost 189 images are of normal and 133 are of abnormal breasts. Promising experimental results have been obtained which depict the efficacy of deep learning for breast cancer detection in mammogram images and further encourage the use of deep learning based modern feature extraction and classification methods in various medical imaging applications especially in breast cancer detection. It is an ongoing research and further developments are being made by optimizing the CNN architecture and also employing pre-trained networks which will hopefully lead to higher accuracy measures. Proper segmentation is mandatory for efficient feature extraction and classification.

Breast cancer detection in mammograms using convolutional neural network

Labeling brain images as healthy or pathological cases is an important procedure for medical diagnosis. Therefore, we proposed a novel image feature, stationary wavelet entropy (SWE), to extract brain image features. Meanwhile, we replaced the feature extraction procedure in state-of-the-art approaches with the proposed SWE. We found the classification performance improved after replacing wavelet entropy (WE), wavelet energy (WN), and discrete wavelet transform (DWT) with the proposed SWE. This proposed SWE is superior to WE, WN, and DWT.

Application of stationary wavelet entropy in pathological brain detection

Automated 3D ultrasound image segmentation to aid breast cancer image interpretation

Photoacoustic image reconstruction method based on compressive sensing

The invention discloses a photoacoustic image optimization method based on the linear delay compensation. The method comprises the following steps that a linear sensor is placed around a target tissue, the target tissue is irradiated through lasers, and the sensor collects photoacoustic signals; the experience sound speed is selected, the size of a pixel is set, different types of delay compensation are selected, the photoacoustic rebuilding is carried out through a delay summation rebuilding algorithm, and the size of delay compensation at difference focusing positions is determined; a curve parameter is determined according to the relation fitting curve of the depth and the delay of a focusing target; the size of delay compensation of each depth is determined according to the rebuilt depth and the fitting curve, and an image is rebuilt through the linear adjusting delay compensation instead of the fixed delay compensation. The linear sensor and the delay summation algorithm are adopted, the linear relation of the delay compensation and the depth is determined through the fixed delay rebuilding, the photoacoustic rebuilding image with the better focusing effect is obtained, and the method has the advantages of being easy to operate and high in imaging quality.

Photoacoustic image optimization method based on linear delay compensation

The invention discloses a multi-mode photoacoustic imaging method combined with limited angle X ray imaging and ultrasonic imaging The multi-mode photoacoustic imaging method includes the following steps that an X ray emitter and a receiver are arranged above and below a target organization respectively, X ray projection data are collected and a limited angle X ray image obtaining the organization is reconstructed; an ultrasonic probe is arranged on one side of the target organization and emits ultrasonic signals to the inside of the organization and reconstructs an ultrasonic image according to the received ultrasonic signals; laser irradiating is carried on one side of the organization to generate the optoacoustic effect, an ultrasonic sensor is used for receiving the signals on the other side of the organization and a photoacoustic image is reconstructed according to the received signals; according to the photoacoustic image, the X ray image and the ultrasonic image are rectified; a multi-mode photoacoustic imaging result is obtained through reconstruction by being combined with the rectified X ray image and the rectified ultrasonic imaging result The features that X ray imaging and ultrasonic imaging are good in imaging quality and high in resolution in individual imaging planes are utilized in the multi-mode photoacoustic imaging method, the image quality of photoacoustic imaging is improved and innovativeness is obtained

Multi-mode photoacoustic imaging method combined with limited angle X ray imaging and ultrasonic imaging

The invention discloses an acoustic velocity correction method for photoacoustic imaging. The method includes the steps of irradiating tissue under test by laser in a photoacoustic imaging system; using a linear sensor to acquire ultrasonic signals emitted by the irradiated tissue; determining a focus evaluation standard, and selecting a focus position; adjusting the acoustic velocity and delay compensation, rebuilding an image by means of delay and sum process, and focusing a selected point; fitting curves of relation between the acoustic velocity and delay compensation according to the relation between the acoustic velocity and delay, and determining curve parameters; performing simultaneous solution on the curves of relation of different focus depths to obtain the corrected acoustic velocity and delay compensation. The linear sensor is used for acquiring signals; the image is rebuilt by means of delay and sum process; a relation equation for the focused acoustic velocity and delay compensation is obtained by adjusting the target acoustic velocities and delay compensations of different positions; the corrected acoustic velocity is acquired by a simultaneous equation; accordingly, operation is simple.

Gu Peng

Papers

Automated 3D ultrasound image segmentation to aid breast cancer image interpretation

Photoacoustic image reconstruction method based on compressive sensing

Photoacoustic image optimization method based on linear delay compensation

Multi-mode photoacoustic imaging method combined with limited angle X ray imaging and ultrasonic imaging

Acoustic velocity correction method for photoacoustic imaging