3rd February 2018
Acacia Nilotica Leaves
Medicine for diabetes?
Diabetes mellitus alters metabolism and leads to hyperglycemia and hyperlipidemia. And these, in turn, lead to problems related to the liver, kidney and blood vessels.
Diabetes has been traditionally treated with Acacia nilotica tree parts. Though many studies demonstrate the anti-diabetic properties of its pods and bark, there is no evidence of the anti-hyperglycemic activity of A. nilotica leaf extract.
Recently, scientists from the University of North Bengal and the Raiganj University, West Bengal collaborated with researchers from the USA, to report that diabetes can indeed be treated with leaves also.
They used animal models with blood glucose levels above 200 mg/dl for their studies. The team used four groups of Type 1 diabetes mice. One was given normal saline. Another group received the anti-diabetic drug, glibenclamide. The other two groups were treated with a low or a high quantity of A. nilotica leaf extracts. The animals were observed for 3 weeks.
The researchers found a reduction in blood glucose level with high quantity A. nilotica leaf extract. The peripheral utilization of glucose also improved.
The leaf extract normalised the hepatic glycogen content and serum insulin level. And lowered the serum creatinine, blood urea nitrogen, and uric acid levels. Moreover, in liver, kidney and skeletal muscles, they found a reduction in the activities of enzymes that lead to diabetes related complications. Therefore, the scientists claim that A. nilotica leaf extracts can be used for treating the complications of diabetes mellitus.
The leaf extract is more cost-effective than conventional drugs. Also, leaves are more easily accessible than pods which only appear seasonally. And harvesting leaves is less harmful to the tree than using the bark as is done in the traditional treatment. These findings will help develop more effective treatments for diabetes.
J. Ethnopharma., 210: 275-286 (2018)
7 January 2018
Early Detection of Breast Cancer
Computer Aided Diagnosis
Breast cancer is a common reason for deaths among women. Early detection through screening techniques has proven key factor to reduce deaths. Though mammography and ultrasound are commonly used screening methods, they have limitations. Biopsy is the confirmatory pathological test for malignancy. However, it is painful.
B. K. Singh and team from the National Institute of Technology, Raipur, now report computer-aided diagnosis for breast cancer. For this, they used two classifiers: back-propagation artificial neural network and support vector machine. The back-propagation artificial neural network is a simple training method even for complex models with thousands of parameters. A support vector machine is a supervised learning method to divide two categories of samples.
The team combined both methods and used data from ultrasound to train the learning machines. Unlike the case with mammography, ultrasound has no radiation or compression. However, Ultrasound images are corrupted by inherent artefacts – speckle noise – leading to low resolution, blurred lesion borders, unclear echo patterns and lack other fine details necessary for diagnosis. So the team used a despeckle filter based on wavelet decomposition to provide smoothening while retaining the details of the image.
They first decomposed the ultrasound images into four sub-bands. The first level decomposition was eliminated and the de-noised image was reconstructed using Inverse Discrete Wavelet Transform. The ultrasound images were of variable sizes. Thus, after speckle noise removal, a region of interest that best enclosed the tumour area was manually extracted in consultation with a radiologist. Thus, a comprehensive set of texture and shape attributes were extracted from the despeckled ultrasound images to differentiate breast tumours.
The scientists used a back-propagation artificial neural network to classify the test samples into malignant or benign. This was considered a first opinion.
The diagnostic results of back-propagation artificial neural networks were then compared with that of the expert opinion. The test samples, for which back-propagation artificial neural network diagnosis matched the expert opinion, were considered correctly diagnosed.
The researchers supplied results of test samples, where back-propagation artificial neural network diagnosis did not match expert diagnosis, to the Support Vector Machine for a second opinion. The diagnosis by Support Vector Machine classifier was considered final.
The team quantitatively demonstrated that a combination of primary and secondary opinions achieves overall improvement in performance. But they feel that use of reports on histology, biopsy and pathology to train the system can improve performance further. Moreover, multimodal imaging – mammography, ultrasound and magnetic resonance imaging – can enhance the accuracy of computer aided diagnosis.
Expert Systems with Applications, 90: 209-223 (2017)
Mary Ranjeetha L M
A new drug in the pipeline
Nano medicines are a major area in nanotechnology. Silver nanoparticles have shown anticancer properties when tested on cancer cell lines. There are various methods for the synthesis of silver nanoparticles. Most are expensive. Moreover, they are often toxic to non-cancerous cells also. Researchers have attempted to synthesise silver nanoparticles using plant extracts and these are reported to bring down costs. These ‘biogenic’ silver nanoparticles seem to retain beneficial properties while reducing toxicity.
Recently, Maidul Hossain and colleagues from West Bengal reported an eco-friendly, inexpensive process for synthesizing silver nanoparticle using an aqueous leaf extract of four leaf clover, Marsilea quadrifolia, a medicinal plant with anti-inflammatory and anti-cancerous properties.
The team reported synthesizing the silver nanoparticle by a simple process of continuous stirring of a mixture of silver nitrate solution with the clover leaf extract. The process does not use reducing and stabilizing agents and is, therefore, not expensive. The synthesis of silver nanoparticles was signalled by a change in colour with maximum absorbance at 435 nm. The nanoparticles exhibited antibacterial activity against E.coli and anticancer activity on cervical cancer cell lines.
The scientists examined the interaction of these silver nanoparticles with human serum albumin and haemoglobin using fluorescent spectroscopy and found no change in the basic structures of these proteins. Though circular dichroism detected a minor change in secondary structure, the researchers feel confident that this silver nanoparticle can be applied safely in the medical ﬁeld.
The study is yet another confirmation of the cost-effectiveness of using plant extracts for nanoparticle synthesis. Given that a large amount of such data has been generated in the recent past, it is time that comparative studies – both metanalysis and experimental – are done to consolidate the field of biogenic nanoparticles for biomedical applications.
J. Mol. Structure, 1141: 584-592