Release date: 2016-11-21
Skin cancer is a high-risk disease in the United States, with 5 million people diagnosed with it every year. The US health system spends $8 billion on this, and more than 100,000 of the 5 million people are skin cancers. The highest rate of melanoma in the world, it kills more than 9,000 lives every year, and this number is still rising year by year.
Looking around the world, melanoma is also a major enemy of public health. In Australia, there are more than 13,000 new melanoma patients each year, and the number of deaths is 1600 per year (the Australian population is very small). In Europe, melanoma kills 20,000 people every year.
In order to curb the rising mortality of melanoma, medical experts need to race against time to diagnose the type of skin cancer in patients as early as possible. The data shows that if the symptoms can be detected in time, the patient's chance of survival in five years is over 98%. However, once cancer cells spread to the patient's lymphatic system, the chances of survival immediately fall below 16%.
The development of technology has provided doctors with very effective tools. Today's professional dermatologists can use early imaging techniques called Dermoscopy to detect melanoma early, but this approach still faces two major challenges.
First of all, the number of professional doctors who can master such technologies in the medical community is seriously inadequate, so patients in many remote areas will be delayed. How to make such diagnostic methods normal and cheap is the biggest challenge in the medical world today.
Secondly, people are not sages, they can't be ignorant, and misdiagnosis in the diagnosis process is a common occurrence. The data show that to diagnose a case of melanoma, at least nine surgical biopsy is required. Surgery biopsy not only causes great physical pain to the patient, but also wastes a lot of money.
Therefore, in response to these two challenges, the medical community needs to intensify innovation to improve the accuracy of melanoma diagnosis. Since the development of human medicine, blood tests have been one of the effective ways, and this method is fast and convenient. The general practitioner can complete the inspection work in a local medical institution or laboratory with simple training. After the test is completed, the data can be an important reference for the attending physician to issue a diagnosis notice. Unfortunately, there is no reliable blood test for melanoma in the medical community today.
Can doctors diagnose skin cancer patients only through pictures in the future?
IBM Research has been working on applying new technologies to skin image analysis. In addition to medical experts, IBM's machine learning, computer vision and cloud computing experts have invested a lot of energy. In addition, the Memorial Sloan Kettering Cancer Center is dedicated to this research.
Recently, IBM has developed a new technology with computer vision. In the future, it is possible to accurately diagnose patients with melanoma by relying on pictures, just as easy as existing blood tests. However, it is not a human medical expert who has such divine power. IBM relies on cloud analysis services. In their vision, the future medical staff only needs to use a dermascope to take a picture of the affected area and upload it to the Cloud Analysis Service Center, and a detailed test report will be obtained shortly. Subsequently, the doctor only needs to analyze the data in the report in detail to determine whether the patient has melanoma. In addition to detecting skin cancer, the data in the report also relates to the patient's subcutaneous cell structure, and the doctor can determine in time whether the patient is still suffering from other diseases.
In 2015, the IBM team released a preliminary study of the diagnostic method. In the report, IBM described the advantages of computer vision diagnostics. With the help of MSK and the International Skin Imaging Partnership, IBM has got the initial data set. In the research, IBM scored a lot, but now the algorithm is still not smart enough, and still needs medical experts to trace the contours of the skin lesions in the picture. As for the accuracy of the diagnosis, IBM said that they have not directly compared with human medical experts (in 2015).
After the release of the research report, IBM has been advancing the advancement of computer vision diagnostic technology, the goal is to allow the computer to automatically identify the skin lesion area while improving the efficiency of the analysis. At the same time, they also invited 8 experts to compare the results of the diagnosis, and found that IBM's computer vision diagnosis method is three times better than the previous diagnosis method, and its accuracy is the same level as 8 industry experts.
The new research report will be officially released in 2017, but Cornell University has already published the report on arXiv's pre-published server.
However, it is important to note that this diagnostic method is still in its early stages. IBM's computer vision diagnostic system is now only involved in the diagnosis of 3,000 skin cancers, and an experienced physician can diagnose 25 patients per day. If you count 20 working days per month, the physician can diagnose 6,000 patients per year. . Therefore, the diagnostic level of the IBM system can only be regarded as a medical student who has just entered the business. In order to put the system into actual combat, IBM has to train more.
However, hope is always there. If the computer vision diagnosis system can take on the heavy responsibility, the doctor can spend more time on the treatment of the disease.
Source: Lei Feng Net
Pumpkin Kernel,Kernel Pumpkin Seed,Squash Seed Kernels,Roasted Pumpkin Seed Kernels
Wuyuan county dafeng oil food co.,ltd , https://www.dafengfood.com.cn