AI is more than ChatGPT, see how cloud computing and AI can help cloud imaging and early cancer screening

1. foreword

   With the recent popularity of ChatGPT, the topic of AI has once again returned to the public's field of vision, and various large models and applications have sprung up like mushrooms after rain. It is true that these applications have helped us improve a lot of work efficiency, but it is always a icing on the cake rather than a timely help.

   The author believes that if AI makes special contributions in the field of life sciences, it will truly benefit mankind. Next, I will share a series of the latest developments and applications of AI in the medical field.

   Imaging is an important part of the medical field and an important means of early cancer screening , so the first article of this article starts with the topic of how AI and cloud computing "empower" traditional medical imaging .

   The basic principle of CT
   

   
   

   

   Since this article is about imaging, we first need to have a general understanding of the basic principles of CT:

   CT can measure the human body with highly sensitive instruments according to the difference in the absorption and transmittance of X-rays by different tissues of the human body, and then input the data obtained from the measurement into the computer. After the computer processes the data, it can be Take a cross-sectional or three-dimensional image of the inspected part of the human body, and find small lesions in any part of the body.

   In layman's terms:

   People can be imagined as tofu, and through the X-ray beam of the human body, we can imagine it as a knife.

   emit x-rays

   X-rays are like cutting tofu, cutting the human body into tiny slices from head to toe.

   attenuation

   During CT scanning, x-rays pass through the cross-section of the human body with a certain thickness, and the intensity of x-rays will attenuate during the process of penetrating the human body.

   Analysis and processing according to different attenuation signals

   CT will have a detector to receive the attenuated x-rays. After a series of complex mathematical operations and computer processing, the attenuated x-ray signals will be converted into images of the cross-section of the human body. Different gray levels represent different tissues and organs , so as to form the CT cross-sectional image we finally see, commonly known as film, film.

   Disadvantages of traditional film

   
   

   

   1: Difficult to carry and organize

   Through the above introduction, we know that there are tens or hundreds of films in one examination, and the radiologist in the hospital will sort out about ten representative films and print them into film.

   If one inspection consists of ten films, a patient with a chronic disease may undergo dozens of continuous inspections. Then there will be hundreds of films, which are difficult to organize and difficult to carry.

   2: Difficult to save

   Due to the influence of the material of the film, it is difficult to preserve the film, especially in a humid and high temperature environment, which will damage the film and make it blurred.

   3: Difficult to find tiny lesions

   We can know from the original introduction of the CT examination above that the more layers of CT, the clearer the image can be seen, and the easier it is to find tiny lesions, while the traditional film is selected by radiologists. layer. If the radiologist fails to find the lesion, the printed film will remain invisible, and neither will subsequent doctors.

   4: Difficult to share for remote consultation

   If in an area with relatively backward medical conditions, it is found through inspection that there may be health problems. When consulting a doctor in a developed area through the APP, the doctor needs to provide an image file. If it is a traditional film, it is difficult to share.

   5: It is difficult to observe the shape of the lesion dynamically

   Some lesions require dynamic observation of morphology, features, magnification, 3D reconstruction, etc., which are difficult to achieve with traditional film.

   Cloud Computing Industry Empowerment

   Solving imaging problems is a problem that both hospitals and patients need to solve, so can the above problems be solved?

   Yes, there are more mature solutions on the market.

   

   

   From the above figure, we can see that cloud manufacturers empower medical institutions and hospitals to upgrade on the image cloud through storage, data, operation and maintenance, cloud computing capabilities, security, and other methods.

   The effect achieved so far

   
   

   • Patients and doctors can query on multiple devices such as APP, browser, and applet

   • Easy to share remote consultation

   • Patients don't have to worry about storage issues for dynamic observation,

   • See the images of each layer and find tiny lesions

   So does the Internet's empowerment of images stop here? No!

   AI empowerment

   

   As shown above:

   Is this millimeter-scale lesion easy to spot visually if the lesion is not marked with a red box?

   If there are a large number of CT scan images (dozens to hundreds) in one examination, it will take a long time for the doctor to make a diagnosis, coupled with a heavy workload, he is prone to fatigue, and manual errors are inevitable.

   Judging from the imaging results, if this is an early-stage lung cancer, if the radiologist misses several layers of the image, or fails to find the lesion, it may have developed into an advanced stage after a few years. This is a very regrettable thing.

   So if we use computer vision plus AI to roughly identify all images and locations that may be lesions, and give prompts to imaging doctors, will it greatly reduce the possibility of missed diagnosis?

   As shown below:

   

   Currently on the market, AI has achieved good results in early cancer screening. The more popular projects are chest CT detection, breast cancer CT screening, and pancreatic cancer screening.

   Summarize
   

   
   
   

   The empowerment of cloud computing is equivalent to breaking through the barriers of imaging information between doctors and patients, and AI will empower radiologists and improve the quality of diagnosis.

   So, is AI's help to the medical field limited to this? How does AI identify early cancers? Welcome to add my WeChat below, continue to follow us, and we will share step by step in the next article.