“Machine learning is the last invention that humanity will ever need to make” – Nick Bostrom
Application in Cancer Research
There are two parts to cancer research, prediction and detection. In the prediction stage, there are three main focuses, which are cancer susceptibility, cancer recurrence, and cancer survivability. Data that the machine will test and train to then predict and diagnose cancer by light absorption, scattering properties, and morphological features. Then, the neural networks help locate the cancer cells by training and testing with key data features leading to cancer detections.
Classifying tumors with the use of machine learning can help cancer researchers better evaluate outcomes based on the growth characteristics of tumors. Recently, IBM created a machine learning algorithm for the prediction of cancer. The machine gave the incorrect treatment for a 65 year old man because IBM used predictive data rather than real information of the patient. This procedure did not work as planned for the IBM engineers.
IBM’s Algorithm
Cancer is the world’s second leading cause of death. There are an estimated amount of 100 different cancerous cells. This causes the prediction of the outcome of cancer to be very difficult. The IBM Research lab in Zurich, has been building a machine learning experimentation to help increase understanding and knowledge of many complex diseases within different cancer types. IBM has been working hard at predicting a disease’s progression through molecular data, which is acquired from diseased tissue samples. This technique will better personalize and design effective treatments for patients.
In their studies, they addressed the task of predicting if a breast cancer survivor will suffer a relapse within a five year period after a patient’s first treatment. Then, the team compared their machine learning algorithm to 14 other similar algorithms that have been previously used for the testing and prediction of breast cancer. Consistently IBM’s machine learning algorithm outperformed all 14 other algorithms. The entire algorithm is open source code, so many new researchers are using their own data to run specific experiments. IBM is looking forward to working with outside partners to translate their machine learning research to benefit doctors and patients.
Conclusion
Machine learning and artificial intelligence is creating a bright future especially in the medical field, which is large and important to society. With the application of machine learning, more insightful observations and better treatment options for patients will soon be possible. Although machine learning and artificial intelligence still have a lot of progress to be utilized within the medical field.
AI business Coding Data Datascientist machinelearning Python technology
Anthony's Blog 