BFBehnam Faraji

Behnam Faraji

*This is an algorithmically generated author record based on publicly available information.

Behnam Faraji has a total of 11 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Simultaneous control of rigidity and hand tremor by adaptive fuzzy Q-learning

    by Behnam Faraji, Saeed Mollahoseini Paghaleh, Meysam Gheisarnejad, Mohammad-Hassan Khooban


  • Advanced non-linear control based on artificial intelligence tuner for hand tremor suppression

    by Behnam Faraji, Korosh Rouhollahi, Akram Nezhadi, Zahra Jamalpoor

    One of the most common symptoms among diverse patients (e.g. Parkinson’s disease, essential tremor) is hand tremor. Deep brain stimulation (DBS) has been a popular method of controlling hand tremor in recent years. Although recent advancements in the DBS system have enabled DBS applications in the context of various neurodegenerative disorders, the deployment of such interface tactics in DBS applications is seriously threatened by un-adaptability concerns and a lack of an accurate model. In response to these issues, the current letter discusses the construction of an intelligent technique for minimizing hand tremor and self-adaptive DBS settings without requiring a proper hand tremor system modeling setup. For this purpose, the hand tremor is suppressed using an intelligent proportional–integral (iPI) as the main controller and the sliding mode (SM) as an observer. Furthermore, as an auxiliary controller, a deep reinforcement learning (DRL) method with the Actor-Critic structure is adaptively incorporated. The feedback parameters of the ultra-local model (ULM) controller will be adaptively modified utilizing the DRL method by online neural network (NN) updating in the proposed strategy. The advantages of the ULM based on DRL structure over state-of-the-art approaches are demonstrated by a complete assessment.


  • A Novel Closed-Loop Deep Brain Stimulation Technique for Parkinson’s Patients Rehabilitation Utilizing Machine Learning

    by Behnam Faraji, Korosh Rouhollahi, Akram Nezhadi, Zahra Jamalpoor

    Deep brain stimulation (DBS) is an effective treatment for movement problems caused by a variety of neurodegenerative illnesses, including Parkinson’s disease (PD). Researchers have created closed-loop DBS (CL-DBS) techniques for reducing hand tremor in many disorders, notably Parkinson’s patients, in recent years. As shown in the studies, hand tremor and rigidity symptoms are two sides of the same coin, with hand tremor leading to rigidity fluctuations and vice versa. The following are the primary contributions of this study: i) depicts the correlation between the two indicators ii) the suggestion of a PPO-based non-integer PID control technique to reduce hand tremor and stiffness at the same time, where the reward function is developed to alter DBS settings when illness levels vary. The proximal policy optimization (PPO) technique is combined with the main controller to provide adaptive learning. The controller parameters of the non-integer PID scheme are regarded as changeable controller coefficients in the proposed strategy, which will be adaptively built by the PPO technique via online learning of its neural networks (NNs). In order to demonstrate the advantages and adaptability of the procedure with different models and patients, the recommended strategy is analysis by computer simulation in a variety of contexts (performance, robustness, and noise) and compared to current methodologies like Integer-Order PID (IOPID) and fractional-order PID (FOPID). The simulation outcomes showed that the created system performed better and more effectively than previous approaches in coping with parameter changes and outside noise, in addition to simultaneously reducing hand tremor and stiffness.


  • Adaptive multi symptoms control of Parkinson's disease by deep reinforcement learning

    by Behnam Faraji, Korosh Rouhollahi, Saeed Mollahoseini Paghaleh, Meysam Gheisarnejad, Mohammad-Hassan Khooban


  • Smart Sensor Control for Rehabilitation in Parkinson's Patients

    by Behnam Faraji, Meysam Gheisarnejad, Zahra Esfahani, Mohammad-Hassan Khooban

    Deep brain stimulation (DBS) is a powerful tool to treat the movement disorders created by Parkinson's tremor. The stimulation of one of the two main parts of basal ganglia (BG) in DBS is often provided without the feedback signal of the tremor which leads to undesirable side effects like cognitive impairment, anxiety, and muscle twitches. In this article, two distinct intelligent controllers are designed to stimulate the parts of a non-linear BG including subthalamic nuclei (STN) and globus pallidus internal (GPi) to overcome the current challenges of the DBS. For this purpose, an interval type-2 fractional-order fuzzy proportional derivative plus integrator (IT2FO-FPD+I) is suggested to control the two parts of BG. The control signals of the IT2FO-FPD+I controller are designed based on the tremor values which are measured by a sensor mounted on the patient's finger. Since the quality of output command of IT2FO-FPD+I controller highly depends on its coefficients, these parameters are adjusted by a sine-cosine algorithm-based wavelet mutation, called SCAWM, in a heuristic scheme. The suggested controller delivers both suitable stimulatory control signals to the brain and decreases the amplitude of tremor impressively. Comparative simulation explorations are performed subsequently to ascertain the superior performance of the suggested structured intelligent controller over that of the state-of-the-art methodologies.


  • An Adaptive ADRC Control for Parkinson’s Patients Using Machine Learning

    by Behnam Faraji, Meysam Gheisarnejad, Maryam Yalsavar, Mohammad-Hassan Khooban

    Parkinson’s disease (PD) is one of the most common diseases that its main complications are hand and head tremors and inflexibility of muscles. One of the prevalent treatments that employ for reducing the symptoms of that is deep brain stimulation (DBS). In practice, a sensor is located in the patient’s finger for detecting and evaluating the tremor values in PD. Using an open-loop control structure for stimulating one area of basal ganglia (BG) is the common approach, but in this work, two areas of BG, named subthalamic nucleus (STN) and globus pallidus internal (GPi) are stimulated in a closed-loop manner separately for i) reducing the intensity of electric field and consequently disappearing the side effects of DBS ii) decreasing hand tremor. In particular, an adaptive Active Disturbance Rejection Control (ADRC) based on a deep deterministic policy gradient (DDPG) and a conventional feedback controller are presented for simultaneous stimulating STN and GPi, respectively. In this way, the control coefficients of the ADRC are considered as the control objective parameters that are designed by the actor and critic neural networks (NNs) of DDPG. The suggested scheme is applied to a BG system model which is frequently studied in the literature. The comprehensive simulation studies are accomplished to confirm the supremacy of the ADRC based DDPG scheme over the state-of-the-art strategies. Moreover, hardware-in-the-loop (HiL) simulations are performed to verify the efficiency of the proposed scheme from real-time perspective.


  • Machine Learning Approach Based on Ultra-Local Model Control for Treating Cancer Pain

    by Behnam Faraji, Meysam Gheisarnejad, Korosh Rouhollahi, Zahra Esfahani, Mohammad Hassan Khooban

    Cancer illness still is one of the most common illnesses in the world, which is constantly rising. Chemotherapy plays a crucial role in treating cancer patients. In this paper, we have presented a novel intelligent sensor for controlling and adjusting chemotherapy parameters which consist of an ultra-local (ULM) controller based on a deep deterministic policy gradient (DDPG). First, the feedback signal is provided using a sensor to calculate the population of cells. Then, a controller sends the proper control commands to the actuator (chemotherapy). In the suggested scheme, the ULM is applied to the dynamic model of cancer. In order to shrink tumor cells and rising immune and normal cells at the same time. Moreover, for improving the performance of the established ULM scheme, a DDPG algorithm with the actor-critic structure is used for tuning the parameters of ULM in an adaptive manner. To demonstrate the supremacy of the DDPG based ULM controller, the conventional ULM and proportional integrator (PI) are also designed for the cancer treatment. Simulation outcomes prove the improved cancer treatment compared to the ULM and PI schemes.


  • A Close Loop Multi-Area Brain Stimulation Control for Parkinson’s Patients Rehabilitation

    by Meysam Gheisarnejad, Behnam Faraji, Zahra Esfahani, Mohammad-Hassan Khooban

    In recent years, deep brain stimulation (DBS) has been one of the most effective methods for treating movement disorders, including hand tremor in advanced Parkinson’s disease (PD). In order to decrees the Parkinson’s tremor, a model of basal ganglia (BG) is often used to design a closed-loop control scheme for DBS. First, a feedback signal is provided using a sensor mounted to the patient’s finger to measure the tremor values. Then, a controller sends the appropriate control commands to the actuator, and finally, the BG areas of the brain actuator are simulated by the actuator. In the present paper, to control Parkinson’s tremor and reduce the value of stimulation intensity efficiently, two areas of BG such as subthalamic nucleus (STN) and globus pallidus internal (GPi) were controlled simultaneously. This strategy provides a reduction in the applied field and side-effects (e.g. muscle contraction and speech disorder) resulting from stimulation intensity. In particular, a new model-free scheme, called intelligent single input interval type-2 fuzzy logic (iSIT2-FL) combined with non-integer sliding mode control (SMC), is proposed to control two areas of BG. In the suggested strategy, an extended state observer (ESO) is established to approximate the unknown BG dynamics, whereas the non-integer SMC is applied to remove the ESO estimation error. Comparative simulation explorations are performed subsequently to ascertain the superior performance of the suggested intelligent control scheme over that of the state-of-the-art approaches.