Shedding Light on Assault

SPEAKER 1: Welcome to Justice Today, the official podcast of the Department of Justice's Office of Justice Programs, where we shine a light on cutting-edge research and practices and offering in-depth look at what we're doing to meet the biggest public safety challenges of our time. Join us as we explore how funded science and technology help us achieve strong communities.

MAYA PILKINGTON: Welcome, everyone. This is Maya Pilkington, Senior Science Writer at the National Institute of Justice. Today, I'm speaking with Dr. Katherine Scafide about her research on bruising supported by NIJ. Dr. Scafide employs her background as a forensic nurse to help those affected by pediatric and adult assault, examining bruising and the impact of skin color, fat, and gender on bruise color. She has been a civil and criminal forensic consultant and currently teaches at George Mason University. She became affectionately known as "The Paintball Lady" for her unique method of creating a bruise, which we will get into in just a moment. Welcome, Dr. Scafide.

DR. KATHERINE SCAFIDE: Thank you. Glad to be here.

MAYA PILKINGTON: Glad to have you here. I'd like to begin first by noting that bruising is one of the common types of soft tissue injury found on victims of violence, whether that's intimate partner violence, sexual assault, child abuse, or vulnerable adult abuse. Documentation of this type of injury can be particularly difficult, and that documentation is extremely important in the criminal justice process. First, could you explain to us what a forensic nurse is, how you typically document assault, and how you capture bruising on the victim?

DR. KATHERINE SCAFIDE: Sure. Well, forensic nurses, which is what I am, are nurses who receive special training, in addition to their nursing training, to provide medical care specifically for patient populations that will likely interact with the legal system, whether it's the criminal justice system or the civil system. As a forensic nurse, we document injuries in a variety of ways. We typically provide a narrative of the injury that we see, in terms of measurements and descriptions. Then we also usually document them on a body map, which is a--a body map is an image of the body, a really generic image, and we indicate where on the body that particular injury might be. And then finally, as best as we can, we always try to get digital images to capture accurately what the injury look like.

MAYA PILKINGTON: Great. When you were on the job as a forensic nurse, you said you encountered a patient that really sparked your interest and trying to expand bruise detection techniques. Could you tell us what happened during that meeting?

DR. KATHERINE SCAFIDE: Sure. I remember clearly back when I was practicing as a forensic nurse, I had a patient who was Black and had very darkly-pigmented skin. And she had reported that she had been sexually assaulted, and we were doing a medical forensic examination in order to assess her well-being and to collect potential evidence. She indicated to me that she had been bitten on her back, and so as a forensic nurse, I do my assessment on that area, I look, I palpate, all based on the history that she's providing me. And I looked and looked and adjusted the angle in which I was looking. I would look again, I would, you know, adjust the lighting and I could not see anything due to the fact that her skin was so heavily pigmented. And unfortunately, as a forensic nurse, it's very difficult for us to document an injury if we can't visually see it. Of course, I collected swabs for potential DNA evidence, I documented where on the body she said she had been bitten, but unfortunately, I had no digital images or no further--or written documentation that I could provide. And if I can't document that particular injury, it doesn't end up being evidence that can really be presented in court to explain what I'm seeing when I do my physical exam.

MAYA PILKINGTON: Uh-hmm. That's right. And so, in 2016, then you turned your experience into a very successful research proposal for NIJ, with a grant that was aimed at more accurately detecting bruising across a variety of skin tones. So how did you set up your research, and what results do you think were the most impactful from that research?

DR. KATHERINE SCAFIDE: Well, since I had that experience practicing as a forensic nurse, I was thrilled to see that NIJ had published a medical protocol for standardizing what a medical forensic exam should involve. And in that protocol, they suggested using alternate light as a method for being able to visualize subtle injuries that were difficult to see. And that's very exciting as a forensic nurse, that we've been able to make--to address this important issue that's creating a bit of a disparity in how we are able to assess injuries on diverse victims. But unfortunately, I didn't see a lot of research evidence to support their recommendation. So, I--it was important for me to put together a proposal for NIJ to address this challenge using rigorous research methods to hopefully provide a scientific foundation for that recommendation. 

So I partnered with Texas A&M University and their team of research nurses and faculty, and we were able to create a randomized controlled trial where we were able to assess bruises over time. And that was really key because bruises go through lots of physiological changes over time. And so it was important for us to know exactly how old that bruise was, but then also examining them repeatedly over time. And each time we assess them, we were comparing white light, which is light that we would typically use in an exam situation, with alternate light, and we tested about 10 different combinations of different wavelengths and colored goggles to see what the best combination might be. 

And so we conducted this study over a couple years. We would assess these particular participants 21 times over four weeks. We also wanted to make sure--because this is where our problem had existed with the current research up to that point, which was mostly Caucasian study participants. We wanted to make sure that we had an incredibly diverse sample because skin pigmentation creates the biggest challenge for assessing bruises on victims of violence. So we used a method called quota sampling to make sure we had equal representation of six different skin color categories. Now, how did we determine what their skin color is? We wanted to make sure we were as objective as possible, so we actually used colorimetry readings. We used a spectrophotometer on the arm and measured what their skin pigmentation was using numerical values, and then you plug them into an equation and that gives you the category for which they fall and it ranges from very light to dark. So we were able to ensure that we had a very diverse sample, which was important. 

And we--in terms of our results, that was even the most exciting part, because after a couple of years of collecting our sample, we were able to discover that violet light and blue light, so we're talking 415 nanometers and 450 nanometers respectively, using a yellow goggles to look at the reflected light were up to five times more effective than white light in detecting evidence of bruising on a diverse sample. So it was quite a relief to see this after studying for a couple of years.

MAYA PILKINGTON: Yeah. And how did you create those bruises?

DR. KATHERINE SCAFIDE: So, we used two different methods, and the reason being is we were trying to reproduce superficial bruises and deeper bruises. So we used a method I know quite well, which is using paintball. I don't know if you've played paintball…

MAYA PILKINGTON: Yes.

DR. KATHERINE SCAFIDE: …but paintball is a very effective method of creating bruises, so. We of course did it under a very standardized method where we were--we shot participants in the upper deltoid eight region, which is the upper arm, behind a barrier we were very safe. And we're able to create superficial bruises that were quite obvious that, you know, that we were able to see right away, but obviously faded over time. 

Alternatively, we also created a second bruise on the forearm of the participant, we randomized the upper and the lower arms when selecting these locations, using a dropped six-ounce steel ball bearing. And this was also another method of creating bruises that was based on other researchers at the time. And the idea is that steel marble would have a longer contact time with the skin surface, and it's a heavier marble traveling at a slower speed, and it would create, theoretically, a deeper bruise on the individual. So we included both bruises in our analysis. A total of 314 bruises.

MAYA PILKINGTON: That's a lot of bruising.

DR. KATHERINE SCAFIDE: Yes, it is.

MAYA PILKINGTON: So how--I mean, after that, how does your method change through time? How has it developed and changed since the beginnings of your experimentation?

DR. KATHERINE SCAFIDE: So the results have been really exciting. And we have actually done deeper analysis in this data cause we have this huge dataset now of bruises and their images. And we have taken a deeper look at how skin color really affects alternate light source use, and we've been able to publish results related to establishing the predictive probability of detecting a bruise based on skin color for certain wavelengths. And so we're able to confirm that on people with very dark skin, you really do need to stick with that violet and blue wavelengths using yellow goggles, which was really fantastic and we're going to be doing more data analysis along the way. 

But this is all great in terms of research, but what does this mean for clinicians? And that's where we really have taken a huge step towards translating that research into practice. So, thanks to NIJ, we got another important grant where we have worked with two forensic nursing units, one unit that does not use alternate light and another unit that does. And we have worked with them collecting data on understanding the context by which they practice. And we use that to develop alternate light source clinical guidelines on how to implement this technology into practice for the purpose of assessing bruises. And they're--they--we have a wonderful acronym, it's called ATLAST. And we're currently working with those two forensic units to finalize these clinical practice guidelines, making sure that they have an input on them. We're also been working with prosecutors and law enforcement, detectives at both locations to have them be involved because these are all stakeholders. And it's important when you implement a change in practice that you think about the stakeholders that would be impacted by it, right? And that includes, obviously in this case, law enforcement and prosecutors. 

So, we're really excited to see how this plays out with these new guidelines. We're going to evaluate the feasibility of them in terms of the impact on the nurse's practice and patient burden with the development. And also--as well as, do we see any change or improvement in the number of injuries being documented? So we're very excited about seeing how this translation into practice project will work out.

MAYA PILKINGTON: Yes. That's a--that's a really big focus for NIJ, is putting this research into practice, so that's going to be fantastic to see. In--are there certain circumstances where you don't recommend this type of screening for individuals?

DR. KATHERINE SCAFIDE: Sure. As part of this project in developing clinical practice guidelines, we did a very thorough search of the literature cause, of course, I'm very familiar with our research, but we want to make sure we really took into account the best available research on this practice. And, at this point in time, it is really not--at least I do not recommend, using alternate light source to assess for injuries on patients that can't give history of physical trauma. 

And the reason for this is that there are a lot of different skin lesions that can cause light absorption. So what we--what we see when we examine the skin using an alternate light source is we're looking for light absorption which looks dark compared to the skin surround it. This is different than fluorescence. Fluorescence, I think, a lot of your--the--your listeners who work in forensics know obviously what fluorescence is. It has a glowing appearance, right? But we're looking for light absorption because bruises and how they break down the different products as they turn into, they absorb light at different wavelengths. And so that's why alternate light works. But also different--other skin lesions in the skin structures that result from those lesions can also absorb light. 

Now, us as nurses, we use a battery of assessment skills when we examine our patients. We--you know, we look, we feel, we ask them a history of what happened and what this--what we're seeing might be. We also, whenever possible, do follow-up exams to see how what we're seeing as--those injuries that we see are healing. So, through that process, we can eliminate a lot of those skin lesions that it might be that we're seeing, you know. We can tell that it's not a tattoo; we can tell it's not a scar; we can tell it's not moles; it's--and that kind of thing, or pressure injuries. But there is still the possibility it might be something else. So that is why a physical history of trauma, it really needs to be present in order for you to be able to determine if this is likely consistent with evidence of bruising, based on the history that you're receiving.

MAYA PILKINGTON: Uh-hmm. Yes, makes sense. Okay. Are there--are there other projects that you have that are in the works to use alternate light sources in forensics with NIJ at the moment?

DR. KATHERINE SCAFIDE: Sure. We're also very excited that we were able to capitalize on that huge dataset that we got from that first project, which included about 26,000 images. And this goes back to my dissertation days where I was really focused on looking at understanding how old a bruise is based on appearance. I mean, this is--this is something that clinicians have been struggling with for decades because it's such a valuable piece of information. If you--if you can tell how old an injury is, it can, you know, really provide validation for a given history of events or possible alleged individuals who might have caused the injury. So trying to figure out ways to do that has really been important, but unfortunately, there hasn't been a lot of science to support it. It's--really, there's just been a reliance on subjective opinion based on clinical experience. 

So I--working with some engineers and some health informatics experts, we have looked at these 26,000 images and determined that it is possible, we feel, to be able to model using artificial intelligence, specifically deep learning, to model how old those bruises are based on the characteristics, you know, the pixels within--the data within the image, and also the other data related to the person who has the bruise. So we're really fortunate that NIJ funded another project for us to--that's a three-year project, that we're hoping will result in really helpful technology for both clinicians and researchers. And at the same time, we're going to create a national repository, at least this is our plan, for bruise images that can then be shared and used as training, and where we can actually install this artificial intelligence once it's finished. So it's very exciting.

MAYA PILKINGTON: That is so exciting. I can't wait to see the results. Katherine, thank you for taking the time to talk with me about your research so that your hard work can be put into practice. That was a fantastic introduction to the use of alternate light for bruise detection. I'm looking forward to hearing more about your current research using artificial intelligence and how it will positively affect victims of assault. 

DR. KATHERINE SCAFIDE: Thank you so much for having me.

MAYA PILKINGTON: This is Maya Pilkington, Senior Science Writer for NIJ. And you've been listening to the Justice Today podcast featuring NIJ grantee, Dr. Katherine Scafide. Thank you, listeners, for joining us and stay tuned for future episodes from NIJ.

SPEAKER 2: To learn more about today's topic or about NIJ, visit the links in the episode description and join us for new episodes every month. 

Opinions or points of view expressed in this episode represent a consensus of the authors and do not necessarily represent the official position or policies of the U.S. Department of Justice. Any products and manufacturers discussed in this episode are presented for informational purposes only and do not constitute produce approval or endorsement by the U.S. Department of Justice.