Getting Started
Welcome back to my blog! I have just completed the first week of my internship at Mayo Clinic. The past few days have been very busy, and I have already learned a lot. I want to share some of my experiences thus far and talk more about my project.
The building itself has a great aesthetic appeal, with its wide construction, spacious corridors, and rich sandstone facade, which effortlessly blends into the surrounding mountains and gently fades into the sunset. A peaceful nature trail surrounds the place, decorated with native desert flora. Upon my entrance, I was cordially greeted by volunteers in the lobby, whose job was to welcome visitors and comfort patients. I immediately sensed a positive atmosphere.
The interior is a colossal maze of hallways connecting a seemingly countless number of rooms. I almost lost my way trying to traverse to the office of Dr. Zwart, the Radiologist who manages the SRP internship program at Mayo Clinic. After receiving a general orientation in the C.O.R.E. bioinformatics lab, I took a brief online course on the Health Insurance Portability and Accountability Act (HIPAA), designed to protect patient information. Patient confidentiality is especially important for my project, as I will be accessing medical records from the hospital database.
I met with my on-site mentor Dr. Silva, who elaborated on the details of my project. Recently, a new type of Hepatocellular Carcinoma (HCC) was identified in pathology, but has yet to be described in radiology. Radiologists do not know what specific image features distinguish this new type of HCC, and clinical imaging tests have completely missed it in the past. This presents a serious concern, as these distinct varieties of HCC require unique treatments. Moreover, HCC is the most common liver cancer, and is the 3rd leading cause of cancer deaths worldwide. As such, the objective of my project is to isolate and characterize this newly discovered type of HCC through the use of texture analysis and machine learning on certain regions of interest gathered from the radiological scans of patients. Hopefully, the results of this research will allow medical professionals to apply more targeted and effective treatments.
After conducting my preliminary research, I began by extracting a dataset of medical images from the Mayo Clinic database. For this, I used a range of softwares, including Illuminate Insight, Vlookup Qreads, Notion, and ImageJ. The process involves filtering specific parameters, anonymizing patient information, and fetching image records, each one identified by a unique accession number. With the help of Sikh, a student at ASU's Barrett's Honors college, I have made significant progress, and expect to have a complete data-set by next week so I can begin texture analysis.
The building itself has a great aesthetic appeal, with its wide construction, spacious corridors, and rich sandstone facade, which effortlessly blends into the surrounding mountains and gently fades into the sunset. A peaceful nature trail surrounds the place, decorated with native desert flora. Upon my entrance, I was cordially greeted by volunteers in the lobby, whose job was to welcome visitors and comfort patients. I immediately sensed a positive atmosphere.
The interior is a colossal maze of hallways connecting a seemingly countless number of rooms. I almost lost my way trying to traverse to the office of Dr. Zwart, the Radiologist who manages the SRP internship program at Mayo Clinic. After receiving a general orientation in the C.O.R.E. bioinformatics lab, I took a brief online course on the Health Insurance Portability and Accountability Act (HIPAA), designed to protect patient information. Patient confidentiality is especially important for my project, as I will be accessing medical records from the hospital database.
I met with my on-site mentor Dr. Silva, who elaborated on the details of my project. Recently, a new type of Hepatocellular Carcinoma (HCC) was identified in pathology, but has yet to be described in radiology. Radiologists do not know what specific image features distinguish this new type of HCC, and clinical imaging tests have completely missed it in the past. This presents a serious concern, as these distinct varieties of HCC require unique treatments. Moreover, HCC is the most common liver cancer, and is the 3rd leading cause of cancer deaths worldwide. As such, the objective of my project is to isolate and characterize this newly discovered type of HCC through the use of texture analysis and machine learning on certain regions of interest gathered from the radiological scans of patients. Hopefully, the results of this research will allow medical professionals to apply more targeted and effective treatments.
After conducting my preliminary research, I began by extracting a dataset of medical images from the Mayo Clinic database. For this, I used a range of softwares, including Illuminate Insight, Vlookup Qreads, Notion, and ImageJ. The process involves filtering specific parameters, anonymizing patient information, and fetching image records, each one identified by a unique accession number. With the help of Sikh, a student at ASU's Barrett's Honors college, I have made significant progress, and expect to have a complete data-set by next week so I can begin texture analysis.
This was an outstanding post. I enjoyed your description of the building a lot. Your project sounds fascinating. How does HCC differ from other liver cancers?
ReplyDeleteHCC is the most common liver cancer, usually diagnosed in patients 50 years or older, and is frequently caused by cirrhosis (long-term damaging and scarring) of liver tissue. It can be attributed to alcohol abuse, auto-immune disease, Hepatitis B or C, and iron overload (hemochromatosis). A prevalent theory is that it originates in the Hepatic stem cells.
DeleteHello! This was very interesting. I really liked the description of Mayo Clinic. :) Its amazing that you are researching a new type of HCC, which hasn't been described yet. Also your engrossment in the field, as I read in your bio, was very moving. I can't wait to see what will happen in the weeks to come!
ReplyDeleteThe pure size of Mayo Clinic seems challenging to navigate. Everything here looks incredibly interesting. From what I know of texture analysis, it would quantify the traits of HCC. Can you further explain what traits you're looking for in HCC? Also, no idea of what machine learning is.
ReplyDeleteAs I currently understand, I will be specifically examining certain nodules in the liver with abnormal growths (sclerosis), and use these features to characterize HCC. I will identify these traits with imaging software, and the process of texture analysis will involve machine learning on the pixel-image data to establish correlations in the image data.
DeleteIt was interesting seeing you discuss what you did during the last week. Your project sounds pretty cool. What distinguishes this new HCC pathologically from other HCCs?
ReplyDeleteHi! This was a fun post to read! I am beginning to have a better understanding of the direction for your project. In the earlier post, you talked a lot about biomarkers. How does biomarkers connect to the texture analysis of HCC? I am looking forward to your next post!
ReplyDeleteThis is a very interesting post. I cant wait to see how your research will help impact patients with HCC. Will you be using different biomarkers to analyze the texture analysis to come up with results?
ReplyDeleteYour research is fascinating. I am interested in how you are going to continue your fascinating research. Can you explain more in depth what isolating is?
ReplyDeleteI really enjoyed this post, and it seems like you are having fun doing this research! What is the main difference/benefit of the new HCC that was discovered?
ReplyDeleteThis project sounds really interesting and I really enjoyed you're description of the building, how was this new HCC identified initially?
ReplyDeleteYour project is very engaging. I really like how passionate and excited you sound. Why is HCC more lethal than other types of liver cancer? Was extracting the dataset a lengthy and time-consuming process? I look forward to your next post!
ReplyDeleteThis comment has been removed by the author.
ReplyDeleteWow, sounds like a lot of fun to be working in such a professional atmosphere. Are the images and data from actual patients that you know of? Looking forward to your update next week.
ReplyDeleteThe images and data are from real patients in the Mayo database. However, in the interest of preserving patient anonymity and following guidelines set by HIPAA, the data will be scrubbed of all identifying information.
DeleteSounds like you are having fun. What are the challenges facing treating a newer type of cancer?
ReplyDeleteI loved this post! Your description of Mayo Clinic is amazing. This post helped me better understand your project. I can't wait to hear more!
ReplyDelete- Jileena