Machine learning models that detect surgical activities in endoscopic videos are instrumental in scaling post-surgical video review tools that help surgeons improve their practice. However, it is unknown how well these models generalize across various surgical techniques practiced at different institutions. In this paper, we examined the possibility of using surgical site information for a more tailored, better-performing model on surgical procedure segmentation. Specifically, we developed an ensemble model consisting of site-specific models, meaning each individual model was trained on videos from a specific surgical site. We showed that the site-specific ensemble model consistently outperforms the state-of-the-art site-agnostic model. Furthermore, by examining the representation of video-frames in the latent space, we corroborated our findings with similarity metrics comparing videos within and across sites. Lastly, we proposed model deployment strategies to manage the introduction of videos from a new site or sites with insufficient data.
According to the CDC, breast cancer is the most common cancer and the second leading cause of cancer related
deaths among women. Metastasis, or the presence of secondary tumors caused by the spread of cancer cells via
the circulatory or lymphatic systems, significantly worsens the prognosis of any breast cancer patient. In this
study, a technique is developed to detect circulating breast cancer cells in human blood using a photoacoustic flow
cytometry method. A Q-switched laser with a 5 ns pulse at 532 nm is used to interrogate thousands of cells with
one pulse as they flow through the beam path. Cells which are pigmented, either naturally or artificially, emit an
ultrasound wave as a result of the photoacoustic (PA) effect. Breast cancer cells are targeted with chromophores
through immunochemistry in order to provide pigment. After which, the device is calibrated to demonstrate a
single-cell detection limit. Cultured breast cancer cells are added to whole blood to reach a biologically relevant
concentration of about 25-45 breast cancer cells per 1 mL of blood. An in vitro photoacoustic flow cytometer is
used to detect and isolate these cells followed by capture with the use of a micromanipulator. This method can
not only be used to determine the disease state of the patient and the response to therapy, it can also be used
for genetic testing and in vitro drug trials since the circulating cell can be captured and studied.
Evanescent field sensing methods are currently used to detect many different types of disease markers and biologically important chemicals such as the HER2 breast cancer receptor. Hinoue et al. used Total Internal Reflection Photoacoustic Spectroscopy (TIRPAS) as a method of using the evanescent field to detect an optically opaque dye at a sample interface. Although their methods were successful at detecting dyes, the results at that time did not show a very practical spectroscopic technique, which was due to the less than typical sensitivity of TIRPAS as a spectroscopy modality given the low power ( ∼ 1 to 2 W) lasers being used. Contrarily, we have used an Nd:YAG laser with a five nanosecond pulse that gives peak power of 1 MW coupled with the TIRPAS system to increase the sensitivity of this technique for biological material sensing. All efforts were focused on the eventual detection of the optically absorbing material, hemozoin, which is created as a byproduct of a malarial infection in blood. We used an optically analogous material, β-hematin, to determine the potential for detection in the TIRPAS system. In addition, four properties which control the sensitivity were investigated to increase understanding about the sensor's function as a biosensing method.
Melanoma is the deadliest form of skin cancer, yet current diagnostic methods are unable to detect early onset of metastatic disease. Patients must wait until macroscopic secondary tumors form before malignancy can be diagnosed and treatment prescribed. Detection of cells that have broken off the original tumor and travel through the blood or lymph system can provide data for diagnosing and monitoring metastatic disease. By irradiating enriched blood samples spiked with cultured melanoma cells with nanosecond duration laser light, we induced photoacoustic responses in the pigmented cells. Thus, we can detect and enumerate melanoma cells in blood samples to demonstrate a paradigm for a photoacoustic flow cytometer. Furthermore, we capture the melanoma cells using microfluidic two phase flow, a technique that separates a continuous flow into alternating microslugs of air and blood cell suspension. Each slug of blood cells is tested for the presence of melanoma. Slugs that are positive for melanoma, indicated by photoacoustic waves, are separated from the cytometer for further purification and isolation of the melanoma cell. In this paper, we evaluate the two phase photoacoustic flow cytometer for its ability to detect and capture metastastic melanoma cells in blood.
Metastasis is a life threatening complex physiological phenomenon that involves the movement of cancer cells from one
organ to another by means of blood and lymph. An understanding about metastasis is extremely important to device
diagnostic systems to detect and monitor its spread within the body. For the first time we report rapid photoacoustic
detection of the induced metastatic melanoma in mice in vitro using photoacoustic flowmetry.
A new photoacoustic flow system is developed, that employs photoacoustic excitation coupled with an ultrasound
transducer capable of determining the presence of individual, induced mouse melanoma cells (B16/F10) within the
circulating system in vitro. Tumor was induced in mice by injecting mouse melanoma cells through tail vein into the
C57BL/6 mice. A luciferase based in vivo bioluminescence imaging is performed to confirm the tumor load and multiple
metastases in the tumor-induced mice. 1ml of blood obtained through cardiac puncture of the induced metastasized mice
was treated to lyse the red blood cells (RBC) and enriched, leaving the induced melanoma in the peripheral blood
mononuclear suspension (PBMC). A photoacoustic flowsystem coupled with an ultrasound transducer is used to detect
the individual circulating metastatic melanoma cells from the enriched cell suspension.
Nanotechnology and the various properties of gold nanoparticles (AuNPs) are quickly changing the field of cancer
detection and treatment. Photoacoustic detection methods show an increase in sensitivity using gold nanoparticle
antibody conjugation, which selectively targets melanoma cancer cells. Instead of targeting melanoma tumors,
we tag single cells, analogous to circulating metastatic melanoma cells. Using an in vitro, stationary cell system
and planar samples, we demonstrate an average of 24% improved optical detectability of melanoma cells tagged
with AuNPs over unprocessed melanoma cells. Tagged cells showed a raised plateau of absorbance from 470nm
to 550nm. Untagged cells showed a general decline in absorption as wavelength increased. The results of our
study have the potential to not only better develop photoacoustic detection of melanoma, but also extend the
viability and use of photoacoustics into detection of otherwise unpigmented cancers.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.