AI Assisted Diagnostic Suite

Christian Burbach
University of Nebraska Lincoln

 

Today, the medical community today faces a significant problem in preventative care

when it comes to diagnosis. Software integration is a significant factor in misdiagnosis. The way physicians communicate with one another is intuitive. However, their software does not communicate in the same communicative way. The process to diagnose a patient may take hours, days or weeks before a final decision may be determined. The length of this process only increases the cost to the hospital and the stress of the patient. In a 2016 Johns Hopkins Study, over 250,000 deaths were attributed to medical error, a rate only behind heart disease and cancer. A significant percentage of that subset were attributed to medical misdiagnosis (McMains, 2016).

The current process starts with the radiologist analyzing their scans and making a decision on what they see on the scan. The scan then moves on to the oncologist, who again examines the scans and decides on a course of treatment. This can take anywhere from minutes to hours, to days, and in some cases even weeks. If an oncologist is not sure of which treatment to provide, they bring the scan to their biweekly tumor board meeting. At said meeting, a team of about 30 medical professionals will discuss the best course of treatment for the patient, a process which varies in length per case, and is by no means expedient.

My solution is to provide an AI-assisted Cancer Diagnostic Suite tool for oncologists to use in tandem with their knowledge and expertise. This AI assisted diagnostic tool would utilize machine learning technology on thousands of CT Scans of various cancers, such as lung, prostate, and breast cancer. Partnered with image recognition modeling software, such as Google’s inception model, the AI would be able to identify tumors immediately based on the scans provided it, all through its data-driven model.

There are many immediate benefits of introducing AI assisted cancer diagnosis to the hospitals nationwide. The quick AI diagnosis stands to save precious time for both the patient and the oncologist. In some cases the tumor may require immediate treatment, the AI-assisted diagnosis suite would provide clear diagnoses drawing upon its national database in milliseconds. The AI would also save hospitals money by increasing productivity through the increase in time available to the physician, who otherwise would have been pouring over his scans, as mentioned before. Most importantly, the AI tool would save lives with its ultra-quick data-driven models, allowing preventative care to take place to keep tumors from metastasizing.

The best preventative care takes place immediately, utilizing all available knowledge and data. Misdiagnosis can delay preventative treatment, or even worse, prescribe treatment counterproductive and detrimental to the health interests of the patient. Hospitals have yet to embrace artificial intelligence programs in their diagnosis procedures fully. The future of preventative health is artificial intelligence assisted diagnosis.



Works Cited

McMains, V. (2016, May 03). Johns Hopkins study suggests medical errors are third-leading cause of death in U.S. Retrieved from https://hub.jhu.edu/2016/05/03/medical-errors-third-leading-cause-of-death/

 

Natalie Stein

Exercise, Fitness & Nutrition Expert | Assistant Professor of Public Health