Clinical applications of endoscopic ultrasound to oncology

Ultrasound segmentation analysis via distinct and completed anatomical borders

PURPOSE

Segmenting ultrasound images is important for precise area and/or volume calculations, ensuring reliable diagnosis and effective treatment evaluation for diseases. Recently, many segmentation methods have been proposed and shown impressive performance. However, currently, there is no deeper understanding of how networks segment target regions or how they define the boundaries. In this paper, we present a new approach that analyzes ultrasound segmentation networks in terms of learned borders because border delimitation is challenging in ultrasound.

METHODS

We propose a way to split the boundaries for ultrasound images into distinct and completed. By exploiting the Grad-CAM of the split borders, we analyze the areas each network pays attention to. Further, we calculate the ratio of correct predictions for distinct and completed borders. We conducted experiments on an in-house leg ultrasound dataset (LEG-3D-US) as well as on two additional public datasets of thyroid, nerves, and one private for prostate.

RESULTS

Quantitatively, the networks exhibit around 10% improvement in handling completed borders compared to distinct borders. Similar to doctors, the network struggles to define the borders in less visible areas. Additionally, the Seg-Grad-CAM analysis underscores how completion uses distinct borders and landmarks, while distinct focuses mainly on the shiny structures. We also observe variations depending on the attention mechanism of each architecture.

CONCLUSION

In this work, we highlight the importance of studying ultrasound borders differently than other modalities such as MRI or CT. We split the borders into distinct and completed, similar to clinicians, and show the quality of the network-learned information for these two types of borders. Additionally, we open-source a 3D leg ultrasound dataset to the community https://github.com/Al3xand1a/segmentation-border-analysis .

Inaccurate pretreatment staging can impact survival in early stage esophageal adenocarcinoma

BACKGROUND

Given the survival advantage of neoadjuvant treatment for locally advanced esophageal cancer, accurate clinical staging is necessary. The aim of this study was to assess the clinical (c) and pathologic (p) staging concordance rates for presumably early stage esophageal adenocarcinoma patients that had upfront esophagectomy (UFE) and evaluate if survival (OS) was negatively affected by inaccurate preoperative staging and subsequent treatment selection.

METHODS

An NCDB retrospective review of nonmetastatic esophageal adenocarcinoma patients that had UFE. The rates of concordance between c and p staging system and OS were calculated.

RESULTS

Of 2775 patients, most patients presented with cN0 (82.8%) and cT1 tumors (53.6%). The overall concordance between c and p staging was 78.8% for T-classification (moderate agreement; weighted κ = 0.729; P < .001) and 78.8% for N-classification (weak agreement; weighted κ = 0.448; P < .001). Patients that were upstaged due to a lack of concordance between T-classification had decreased 5- and 10-year OS (30% and 16%, P < .001) and those upstaged due to discordant N-classification had decreased 5- and 10-year OS (28% and 23%, P < .001)."

CONCLUSIONS

Preoperative staging of esophageal adenocarcinoma has moderate reliability and accuracy for predicting pT and pN classification. Up to 25% of patients have discordant clinical and pathological staging, which impacts OS.

Endoscopic ultrasound in oncology: An update of clinical applications in the gastrointestinal tract

An accurate staging is necessary to select the best treatment and evaluate prognosis in oncology. Staging usually begins with noninvasive imaging such as computed tomography, magnetic resonance imaging or positron emission tomography. In the absence of distant metastases, endoscopic ultrasound plays an important role in the diagnosis and staging of gastrointestinal tumors, being the most accurate modality for local-regional staging. Its use for tumor and nodal involvement in pre-surgical evaluation has proven to reduce unnecessary surgeries. The aim of this article is to review the current role of endoscopic ultrasound in the diagnosis and staging of esophageal, gastric and colorectal cancer.

Adenocarcinoma of Small Bowel

Small bowel cancer is one of the rarest cancers in the gastrointestinal tract. The diagnosis is usually late and most patients presented with the advanced stage. Because of this rarity, there is limited data when making decisions for treatment and biological behavior. Most forms of the cancer occur in the duodenum with surgery being the treatment of choice if the cancer is operable. Chemotherapy has an accepted role in duodenal cancer, with the best form being regimen, which yields the best result in combination with capecitabin and oxaliplatin. Our case patient was present with liver metastasis and a huge mass in her first duodenal region so we were required to use chemotherapy and radiotherapy. Like other duodenal cancers, the metastasis decreased her survival and she died about 13 months after diagnosis.

EUS for choosing best endoscopic treatment of mesenchymal tumors of upper gastrointestinal tract

AIM: To evaluate the value of endoscopic ultrasonography (EUS) in the choice of endoscopic therapy strategies for mesenchymal tumors of the upper gastrointestinal tract.

METHODS: From July 2004 to September 2010, 1050 patients with upper gastrointestinal mesenchymal tumors (GIMTs) were diagnosed using EUS. Among them, 201 patients underwent different endoscopic therapies based on the deriving layers, growth patterns and lesion sizes.

RESULTS: Using EUS, we found 543 leiomyomas and 507 stromal tumors. One hundred and thirty-three leiomyomas and 24 stromal tumors were treated by snare electrosection, 6 leiomyomas and 20 stromal tumors were treated by endoloop, 10 stromal tumors were treated by endoscopic mucosal resection and 8 stromal tumors were treated by endoscopic submucosal dissection. Complete resection of the lesion was achieved in all cases. Of the mesenchymal tumors, 90.38% diagnosed by EUS were also identified by pathohistology. All wounds were closed up nicely and no recurrence was found in the follow-up after 2 mo.

CONCLUSION: EUS is an effective means of diagnosis for upper GIMTs and is an important tool in choosing the endoscopic therapy for GIMTs, by which the lesions can be treated safely and effectively.

Endoscopic ultrasonography guided celiac plexus neurolysis and celiac plexus block in the management of pain due to pancreatic cancer and chronic pancreatitis

Pain is a common symptom of pancreatic disease and is frequently difficult to manage. Pain relief provided by narcotics is often suboptimal and is associated with significant side effects. An alternative approach to pain management in pancreatic disease is the use of celiac plexus block (CPB) or neurolysis (CPN). Originally performed by anesthesiologists and radiologists via a posterior approach, recent advances in endoscopic ultrasonography (EUS) have made this technique an attractive alternative. EUS guided celiac plexus block/neurolysis is simple to perform and avoids serious complications such as paraplegia or pneumothorax that are associated with the posterior approach. EUS guided CPN should be considered first line therapy in patients with pain due to pancreatic cancer. It provides superior pain control compared to traditional management with narcotics. A trend for improved survival in pancreatic cancer patients treated with CPN has been reported, but larger studies are needed to confirm this finding. At this time, the use of EUS guided CPB cannot be recommended as routine therapy for pain in chronic pancreatitis since only one-half of the patients experience pain reduction and the beneficial effect tends to be short lived. EUS guided CPB and CPN should be used as part of a multidisciplinary team approach for pain management.