This Small Business Innovation Research (SBIR) Phase II project seeks to improve medical image analysis and reporting. The project proposes research and innovation on quality of diagnosis, richness of knowledge exchange, privacy and scalability of merging analysis and reporting of medical imaging for the detection of breast cancer. The company proposes a system which allows radiologists to focus on analyzing the image while the system automatically produces a report in the background. The reports are image-based, provide audience-specific levels of detail, and are backwards compatible to text-based reports. The SBIR Phase I project established the efficiency improvements that the proposed system can provide. In Phase II, this project will investigate cloud computing as a model for operating and delivering the company's mammography integrated analysis and reporting software. Cloud computing has proven to be a scalable and cost-effective means of delivering software services to customers and will thus be an important capability of health-care IT providers. However, the image storage and security requirements for medical images will necessitate improvements over existing cloud solution architecture. The Phase II project will develop new methods for delivering richer image-based reports and robust security via the cloud.
Breast cancer affects nearly one in eight women, at a substantial economic and societal cost. Mammography screening is the most effective way to detect breast cancer in its earliest and most treatable phases. However, there are many inefficiencies and quality problems in screening mammograms. While a shortness of radiologists specializing in mammography is driving up costs, there are significant false-negative rate and false-positive rates in screening mammography. This situation creates a clear opportunity for innovations that can provide increased efficiency while maintaining or improving quality. Changes in health-care legislation are yielding incentives for health-care facilities to transition to all-electronic systems. As they do so, many health-care facilities are embracing cloud-based delivery and Software-as-a-Service models. If successful, the Phase II project will continue with research and innovation for the delivery of an integrated analysis and reporting software that will improve the quality of medical image diagnosis while decreasing costs.
MEDKEN, a medical knowledge software and service company, has developed a system to improve the accuracy and efficiency of the radiology diagnosis process. The current radiology report generation process is rife with unnecessary practices and procedures. For example, in current practice, the act of examining the images and generating the final diagnostic report are two discrete tasks, conducted with two different modalities, an image exploration system and text-based reporting system. This is an inefficient process which causes information loss and in some cases, diagnostic error. MEDKEN has developed a software solution, MEDKEN Touch, which seamlessly merges radiology reporting with the image analysis and interpretation process. Instead of examining the electronic images on one monitor and turning to a separate monitor to write the report, the radiologist, using MEDKEN Touch, is able to keep her eyes on the images. MEDKEN Touch is an annotation-based reporting system which is an overlay to the image viewer. The radiologist focuses on identifying and annotating abnormalities in the tissue, while the report is automatically generated in the background. By eliminating the clerical side of the reporting process, MEDKEN Touch improves efficiency and allows the radiologist to review more studies in less time. MEDKEN Touch also helps to eliminate reporting errors by automatically generating the anatomical location statement of each lesion based on the radiologist's annotations. An early prototype of MEDKEN Touch was used in a clinical study at the University of Chicago Medical Center to examine the effects of automated annotation-based reporting on mammography workflow. The research team was able to show an average of 21% improvement in efficiency. That is, radiologists were able to analyze and report on 21% more studies in the same amount of time using MEDKEN Touch. Since the initial study, the MEDKEN engineering team has transformed the software from a mammography-specific prototype into a fully-functional multi-modality reporting system compatible with all study types. MEDKEN Touch has been demonstrated at several industry shows, including the Radiological Society of North America (RSNA) annual conferences of 2011 and 2012, the 2013 Society of Breast Imaging conference, and the 2012 Consumer Electronics Show. Reception has been overwhelmingly positive. The need for a more efficient and accurate reporting workflow tool in radiology has been recognized by many of the larger companies in the radiology software industry. Several of these companies have approached MEDKEN with a desire to integrate the automated reporting of MEDKEN Touch with their existing medical image viewing software products. The MEDKEN engineering team is nearing completion of a software development kit (SDK) version of Touch which can be easily integrated with third-party image viewing applications. This will help to expand the commercial market opportunity for MEDKEN and therefore broaden the impact of MEDKEN's improvements to the quality and efficient of radiology reporting.
