Week 2 - Juan Boza: Immersing in Dermatology

Immersing in Dermatology!

The second week of immersion has finished - Time flies when you are having fun! Well, fun is usually associated with happiness, which does not apply to this particular situation. Here, fun refers to joy -  the joy of learning, observing, and helping. To start, let me just tell you that a week in dermatology is quite an experience for many, many reasons. Dermatology is one of the most difficult specializations. All outer surface has to be examined for lesions, moles, abnormalities.

   Fig 1. A. Common features of a subungual hematoma. Feat-
   ures are usually observed under the Dermatoscope (Image
   taken from Mun et al.). B. Image of a left toenail with Hema-
   toma.
The learning curve this week was steep. I learned about different nail conditions, including Onychomycosis, Psoriasis, Hematoma, Melanonychia, Lichen Planus, Paronychia, etc. For general understanding: Onycho - nail (fingernail or toenail). For the learning this week, let's focus on Hematoma. Subungual hematomas or hemorrhages are collections of blood in between the nail bed and the nail plate. There are several features for this condition, including periungual hemorrhages, satellite globules, peripheral fadings, homogeneous patterns, and streaks (Fig. 1A). Note that not all these features have to be present. To observe these features, dermatologists use a device called "the Dermatoscope" - a device with magnification for nail and skin examination. The cause of hematomas is usually associated with trauma - like closing the car's door on your hand or dropping heavy objects on your feet. Probably you are thinking, "Why would I care about a traumatic event?" - Apart from the obvious pain-related aspects, hematomas sometimes are confused with melanonychia or other conditions, which are usually related to melanomas. WOW right?, I bet you did not know that you can get melanoma under your nail.  For this, dermatologists must take a look at hematomas, especially when the cause is unknown. An example of this condition is presented in Fig. 1B. This shows a "textbook" case of a hematoma, where the bruises and blood are well defined. 

Fig 2. A. Original image along with features extracted from
color channels. B. Superposed features on channel images
for visualization of features.

As you may be thinking, observation is key in the diagnosis of these conditions! So, you might be thinking that it is kind of unreasonable for a patient to come all the way to the clinic for a 10-minute appointment. Spending money and time just so the dermatologist can quickly look and determine what is underneath the nail. To avoid this, I mentioned in my first post that I am trying to use artificial intelligence (AI). In a situation like this, the AI can be used to analyze the images provided by the patient to determine if the patient should come to the clinic (or do a video visit). Furthermore, the AI provides a preliminary diagnosis, which can be used to rule whether a biopsy is necessary. While the AI is not ready to be used yet, I developed a preprocessing algorithm for feature extraction this week. As shown in Fig. 2A, the algorithm extracts sharp changes from the images. The goal is to feed the AI the original image along with these features to increase its efficiency. Fig. 2B shows the superposition of these features on the respective channel. An important aspect of these features includes better visualization of color fading on the red channel, an aspect of hematoma. 

And well, I don't know how your week was, but mine was amazing! 
Stay tuned for more...

Reference:
Mun, J‐H., et al. "Dermoscopy of subungual haemorrhage: its usefulness in differential diagnosis from nail‐unit melanoma." British Journal of Dermatology 168.6 (2013): 1224-1229.



Comments

Post a Comment

Popular posts from this blog

Week 1

For my first week of immersion I attended the OR's of a few different physicians as my clinical mentor Dr. Jason Spector of plastic surgery and reconstruction was on vacation this week. I was able to see three different hernia repairs, and one double mastectomy and bilateral DIEP flap.  I think my favorite thing I saw was the double mastectomy and bilateral DIEP flap. From what I understand, this is a procedure in which both breasts were removed and fat from the stomach is used to reconstruct the breasts. The novelty of the DIEP flap is that it does not require the removal of muscle which allows patients to recover quicker. Additionally, blood vessels from the fat in the stomach are left in tact and "hooked up" to corresponding vessels in the breast. All in all, the procedure essentially amounted to a breast reconstruction and tummy tuck. It was a 10 hour surgery that required multiple doctors and departments, as this was not just a purely cosmetic procedure- the patient ...
Read more

Week 1

Image
  During my first week of Immersion at Weill Cornell Medicine, my clinical mentor, Dr. Jason Spector, was out of town. As a result, I spent my time shadowing other clinicians, familiarizing myself with the layout of the hospital, and introducing myself to his lab.  I witnessed four major types of surgery: gastric bypass, breast reconstruction, hernia repair, and craniectomy. Gastric bypass surgeries reduce the working size of the stomach with the intention of helping the patient lose weight by reducing appetite. This is accomplished by creating a small pouch from the stomach and passing over the first section of the small intestine.  The first breast reconstruction surgery was for a patient who was undergoing radiotherapy for breast cancer. Temporary inflatable implants were surgically inserted as placeholders while she undergoes therapy. After her treatment is complete, she will return for polymeric-based implants. This case was interesting to witness since...
Read more

Week 1 - Robert Hawkins

Research - My project for this summer is going to focus on using a large national database that contains blood/plasma biomarker data, cognitive scores, and image data on ~6000 subjects that are either healthy, have mild cognitive impairment, or have Alzheimer's disease. Specifically, I will be looking for associations between neutrophil activation and reductions in cerebral blood flow, which is what my (Schaffer-Nishimura) lab in Ithaca is also investigating. I like this project because in the Schaffer-Nishimura Lab we only work with mice, and so it nice to now be able translate that work and to analyze some of the same measurements in human subjects. I am also learning how to use the statistical program 'R' for this project, which I think is a good skill to develop. Much of my week has been spent looking at this Alzheimer database and trying to find/extract the data that I need. I've already run into a few problems with this dataset in that there are not many patients ...
Read more